bioelectrochemical systems for energy ...―wastewater treatment process‖. australian provisional...

BIOELECTROCHEMICAL SYSTEMS FOR

ENERGY RECOVERY FROM WASTEWATER

KA YU CHENG

Bioelectrochemical Systems

for Energy Recovery from Wastewater

Ka Yu Cheng

BSc (Hons); M.Phil.

A thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in

Environmental Engineering

Faculty of Sustainability, Environmental and Life Sciences

Murdoch University

WA, Australia

November 2009

i

Abstract

The global concerns of climate change and energy crisis have provoked research efforts

to develop energy-efficient alternatives to conventional activated sludge wastewater treatment

processes. Recently, bio-electrochemical systems (BES) such as microbial fuel cells (MFCs) and

microbial electrolysis cells (MECs) have emerged as a promising technology for simultaneous

energy recovery and wastewater treatment. These systems harness the capacity of

microorganisms for the catalysis of electrochemical reaction. In MFCs, chemical energy in the

form of organic compounds in wastewater is directly converted into electricity. While in MECs,

external electricity is provided to enable more valuable products (e.g. hydrogen) to form at the

cathode.

Over the past decade, our knowledge on BES is gaining momentum. However,

wastewater treatment using BES has not been successful on an industrial-scale. Many

technological bottlenecks still remain unresolved and our understanding of microbe-electrode

interactions in BES is incomplete. The overall aim of this thesis is to generate understanding that

will be helpful in the development of an energy recovering wastewater treatment process using

BESs. The scope of this thesis comprises two themes. The first theme is to study the

fundamental aspect of BESs with an emphasis on microbe-anode interactions, the reaction that

makes it possible to use organic waste substances as the electron donor for BES. The second

theme is to quantify rate limiting steps and to develop practical solutions to overcome the

established bottlenecks making room for the development of new technologies of BES for

wastewater treatment and related purposes.

Most of the experiments were conducted using a sub-liter scale two-chamber BES

equipped with a cation exchange membrane. An electrochemically active biofilm was

established at the anode (graphite granules) from an activated sludge inoculum. A synthetic

wastewater with acetate as the sole electron donor was used throughout the study. The cathodic

electron acceptor was either potassium ferricyanide (K3Fe(CN)6) or dissolved oxygen. No

chemical catalyst (e.g. platinum) was applied to the electrodes. In some experiments, the BES

was coupled with a potentiostat to precisely control the biofilm-electrode potential or to perform

voltammetric analysis.

The results showed that activated sludge bacteria could readily initiate a highly effective

anodophilic biofilm (415 W·m-3

after five days with a Fe(CN)63+

cathode), provided that factors

such as electrolyte pH, external resistance and cathodic oxidizing power were not limiting. From

the Coulombic efficiency of over 80% the microbial activity could be recorded by online

monitoring of the current. This allowed a detailed study of the affinity for the anode of biofilm.

In analogy to the well known Michaelis-Menten kinetics, a half-saturation anodic potential (kAP)

was established at which the microbial metabolic rate reached half its maximum rate. This kAP

value was about -455 mV (vs. Ag/AgCl) for our acetate-driven biofilm. A critical anodic

potential (APcrit.) of about -420 mV (vs. Ag/AgCl) was defined that characterizes both the

bacterial saturation by the electron-accepting system and the maximal MFC power output. This

information is useful for MFC modeling and optimization.

Although online process control is used for many bioprocesses it is not established for

MFC. A new approach was developed that enabled voltammetric studies of the biofilm-anode,

ii

without using potentiostats but by feedback controlling a variable external resistance of the

running MFC. This approach could perform the conventional cyclic voltammetry of a MFC

without interrupting its operation.

In MFC the anodic reaction is proton liberating while the cathodic reaction is proton

consuming. This leads to perhaps the major limitation in MFC operation known as a pH gradient.

This limitation was addressed by using a novel operational regime: the intermittent polarity

inversion. At electrode potential of -300 mV (vs. Ag/AgCl) the alternating supply of acetate and

oxygen to the biofilm resulted in the generation of an anodic and cathodic current of +240 and -

80 mA (+1500 and -500 A·m-3

), respectively. Since the anodic reaction is proton-liberating

while the cathodic reaction is proton-consuming, such operational regime prevented the

detrimental build-up of a pH gradient enabling a prolonged operation of the MFC without using

costly pH control methods (dosing of acid/base or chemical buffers).

The intermittent polarity inversion showed signs of the ―anodophilic bacteria‖ being able

to catalyze not only the anodic oxidation of acetate but also the cathodic oxygen reduction. The

presence of ―anodophilic bacteria‖ at the cathode could enable a 5-fold increase of power output

(from 5.6 to 27 W·m-3

). This is the first evidence that a BES biofilm can catalyze both the

forward and backward electron flow with a single electrode. Based on this finding, a novel

scalable, membrane-less BES configuration, termed rotatable bio-electrochemical contactor

(RBEC) has been developed.

Similar to rotating biological contactors (RBC), the RBEC consists of a cylindrical

water-holding vessel (ca. 3 L) which houses an array of carbon cloth coated discs (electrodes)

mounted onto a central horizontal rotatable shaft. Each disc consists of a water-immersed anodic

and an air-exposed cathodic half connected via a resistor. No ion-exchange membrane and

wastewater recirculation were required as the air-water interface separated anode from cathode.

A polarity inversion aiming at overcoming the pH gradient and enable the biofilm to catalyze

both the anodic and cathodic reaction could be obtained by merely turning the disc array a half

turn.

An electron flow from the submersed half disc to the air exposed half disc established

with the moisture film on the air exposed cathode allowing the ionic charge transfer. As with

other MFC any measured current could be documented to be linked to COD oxidation. The

COD removal caused by the action of the intermittently turning discs was increased by about

30% by merely allowing an electron flow between the anodic and cathodic disc halves. This

result suggests that the treatment performance of traditional RBC may be significantly increased

by using suitable conductive material as the discs.

By raising the cathodic potential from about -500 to -1200 mV (vs. Ag/AgCl) using a

potentiostat the cathodic limitation could be alleviated allowing an increase in electron flow and

COD removal rate to 1.32 kg COD m-3

day-1

(hydraulic retention time 5h). While the COD

removal rate was comparable to that of an activated sludge system, the potentiostatically

supported RBEC removed COD more energy-efficiently than activated sludge systems (0.47 vs.

0.7-2.0 kWh kgCOD-1

), even though it was not optimized. The RBEC could also enable

electrochemically-driven hydrogen gas or methane gas production when operated as a MEC

under fully anaerobic condition.

iii

Overall, this thesis has extended our understanding on how electrochemically active

microorganisms behave in BES. Especially, the discovery of the bidirectional microbial electron

transfer property may shed light not only on BES development, but also on the context of

fundamental microbiology. The new RBEC configuration may widen the functionality or

suitability of BES for a large-scale wastewater treatment application. Nevertheless, further

process optimization is needed. In particular, the cathodic reaction still remains as the key

process bottleneck. Future efforts should thus be oriented towards improving and elucidating in

details the mechanisms of the microbe-cathode electron transfers.

iv

Declaration

I hereby declare that this submission is my own work and that, to the best of my

knowledge, it contains no material previously published or written by another person nor

material which to a substantial extent has been accepted for the award of any other degree or

diploma of the university or other institute of higher learning, except where due

acknowledgment has been made in the text.

Ka Yu Cheng

-----------------------------------------------------

(Signature)

Items derived from this thesis:

1. Patent

―Wastewater Treatment Process‖. Australian Provisional Patent (filed in July 2009). Application number

2009903544. (Chapter 7 and 8)

2. Publications

K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2007) Limitations of bio-hydrogen production by anaerobic

fermentation process: an overview. American Institute of Physics Conf. Proc. Vol. 941, 264-269. (Cited in

Chapter 1)

K. Y. Cheng, G. Ho and R. Cord-Ruwisch (2008) Affinity of microbial fuel cell biofilm for the anodic potential.

Environmental Science and Technology. Vol. 42(10), 3828-3834. (Chapter 3)

K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2009) A new approach for in situ cyclic voltammetry of a

microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry. Vol. 74, 227-231 (Chapter 4)

K. Y. Cheng, G. Ho and R. Cord-Ruwisch. (2010) Anodophilic biofilm catalyzes cathodic oxygen reduction.

Environmental Science and Technology. Vol. 44(1), 518-525. (Chapter 6)

K. Y. Cheng, G. Ho and R. Cord-Ruwisch. Rotatable bio-electrochemical contactor (RBEC)-A novel

wastewater treatment hybrid technology of rotating biological contactor and bioelectrochemical system.

To be Submitted. (Chapter 7)

K. Y. Cheng, G. Ho and R. Cord-Ruwisch. A rotatable bio-electrochemical contactor (RBEC) enables

electrochemically driven methanogenesis. To be Submitted. (Chapter 8)

3. Award

HUBER Technology Prize 2008 (first prize). In IFAT 2008 - 15th International Trade Fair for Water - Sewage

- Refuse - Recycling, Munich, Germany. (Part of Chapter 2)

http://pericles.ipaustralia.gov.au/ols/auspat/applicationDetails.do?applicationNo=2009903544

v

Acknowledgments

I wish to express my gratitude to a number of individuals. Without their support and help

over the past three years, my PhD journey would not be as meaningful and unforgettable as it is.

First of all, my thanks should go to my PhD supervisors who have given me the greatest

freedom ever in steering the research direction throughout the thesis. I would like to express my

deepest gratitude to each of them:

o Prof. Goen Ho — who has constantly offered me opportunities, supports, guidance,

advices and encouragement… Thanks Goen!

o Dr. Ralf Cord-Ruwisch — who has changed my view of dealing with sciences and has

offered me his unlimited inspirations on both sciences and philosophies…,. ―Suck &

See!‖ I am also thankful for the opportunities to participate with his various research

and teaching activities, also for his encouragement, coffee and fun times … Thanks Ralf!

Now, I would like to express my immense gratitude to:

The examiners of this thesis for their valuable expert comments:

o Prof. Lars Angenent, Department of Biological and Environmental Engineering, Cornell

University, USA;

o Prof. Sang-Eun Oh, Department of Biological Environment, Kangwon National

University, South Korea; and

o Prof. Uwe Schroder, Sustainable Chemistry & Energy Research, Institute of Ecological

Chemistry, Technical University-Braunschweig, Germany;

Murdoch University for offering me the full scholarship and living allowance. Thanks

Goen and Ralf again for topping up my scholarship. Without these financial supports I

would not be able to thrive through my research journey in such a beautiful and peaceful

country.

Commericialization office of Murdoch Univeristy for funding (Discovers Grant) the

works in Chapter 7 and 8 of this thesis. Thanks Dr. Patty Washer and Ms. Sam Dymond

of this office for their expert supports on the patent application in the later phase of the

thesis.

My office and laboratory colleagues for their encouragement and precious friendship —

Isabella, Davina, Nora, Lee Walker, Wipa, Donny, Mitch, Suwat, Ying, Raj, Liang and

Chia …Thanks Mates!

All technical staff from the university mechanical workshop. Special thank is given to

Mr. Fritz Wagen for his brilliant contribution to the design and manufacture of the RBEC

reactors. Also, thanks Fritz for showing me your amazing homemade little Jet-engine in

your own backyard workshop!

vi

Mr. John Snowball from the electronic workshop for his excellent assistance on the

manufacture of various electronic devices throughout the project. His valuable technical

advice is also highly appreciated.

Dr. Lucy Skillman for her technical assistance with the biofilm analysis and Mr. Gordon

for his technical assistance on using the scanning electron microscope.

Dr. Korneel Rabaey from the Advanced Water Management Centre, The University of

Queensland for his kind assistance in manufacturing the high quality bioelectrochemical

reactor (and the accessories such as the graphite granules and the cation exchange

membrane) used in this thesis and also for his valuable advice on the topic.

Staff in the School of Environmental Science and the School of Biotechnology and

Biological Sciences for creating such a friendly, dynamic and harmonic environment to

work in and also for allowing me to use various facilities and equipments.

Prof. Jonathan Wong (Hong Kong Baptist University), my honors and masters supervisor

who has constantly given me encouragement and advice since the first day of my

research journey.

Last but not the least; my deepest appreciation are given to my wonderful family: my parents,

my brothers (Karson and Kelvin), my wife Suki and her family. – Thanks for your

understanding, encouragement in every part, continuous support & love…

Finally I wish to share the joy of completing this work with all of the aforementioned

individuals, without their ‗source of sustainable energy‘, completing the thesis alone would be

highly ―unfeasible‖!

Thank You!

Ka Yu

Bateman - Perth

Winter 2009

vii

Table of Contents

Abstract i Declaration iii Acknowledgements v Table of Contents vii

Page

Chapter 1 Introduction and Aim 1 1.1 Background 2

1.1.1 We need to reduce Global Energy Consumption 2

1.1.2 Wastewater Treatment: From "Energy-to-Waste" to "Energy-from-Waste" 3

1.1.2.1 Activated Sludge Processes: Energy-to-Wastewater 4

1.1.2.1.1 Bringing Oxygen from Air to Wastewater requires Energy 4

1.1.2.1.2 Electron Transfer instead of Mass Transfer 4

1.1.2.2 Options for Wastewater-to-Energy 6

1.1.2.2.1 Methanogenic Anaerobic Digestion 6

1.1.2.2.2 Fermentative Hydrogen Production 7

1.1.3 Bioelectrochemical Systems for Sustainable Wastewater-to-Energy 9

1.1.4 Overview of Bioelectrochemical Systems 10

1.1.4.1 BES: A Fast Growing Field of Research 10

1.1.4.2 Basic Features of BES 10

1.1.4.3 Principles and Thermodynamics of Bioelectrochemical Conversion in BESs 12

1.1.4.3.1 MFC and MEC: Similarities and Differences 14

1.1.4.3.1.1 Similarities: Anodic Oxidation is catalyzed by Microbes 14

1.1.4.3.1.1.1 Dilemma: A Lower or a Higher Anodic Potential? 15

1.1.4.3.1.2 Difference: Exergonic or Endothermic Cathodic Reduction? 16

1.1.4.4 Microbial Oxidation of Organics Using an Insoluble Electrode as Electron Acceptor

17

1.1.4.4.1 Exocellular Electron Transfer and Mediators 18

1.1.4.4.1.1 Self-Mediated Exocellular Electron Transfer 19

1.1.5 Limitations in BES Processes 20

1.1.5.1 The Use of Membrane Separators leads to pH Splitting Phenomenon 20

1.1.5.1.1 pH Splitting reduces BES Performance 21

1.1.5.1.1.1 A Fundamental Understanding on the Effect of pH Change on BES Performance

21

1.1.5.1.2 Conventional and State-of-the-Art Approaches for pH Control in BES 23

1.1.5.2 Poor Cathodic Oxygen Reduction in MFC 24

1.2 Aim and Scope of the Thesis 25

Chapter 2 Establishing an Anodophilic Biofilm in a MFC with a Ferricyanide-Cathode and Online pH Control

27

2.1 Introduction 28

2.2 Experimental Section 30

2.2.1 Microbial Fuel Cell Construction 30

2.2.2 Bacterial Inoculum and Medium 31

2.2.3 Start-Up and Operation of a Ferricyanide-Cathode MFC 32

2.2.4 Start-up of Anodophilic Activity in a Potentiostatic-Coupled MFC 33 2.2.5 Performance of a Dissolved Oxygen-based Catholyte 34

2.2.5.1 Effect of Catholyte pH and Phosphate Buffer Concentration 34

2.2.6 Calculation and Analysis 36

2.2.6.1 Determination of Voltage, Current and Power Generation 36

2.2.6.2 Polarization Curve Analysis 36

2.2.6.3 Acetate Analysis 36

2.3 Results and Discussion 38

viii

2.3.1 Quick Start-up of MFC using Ferricyanide-Cathode and Continuous pH-Static Control

38

2.3.1.1 The Cathodic Oxidation Power could influence the Anodic Microbial Activity only at a Low External Resistance

40

2.3.2 Potentiostatically-Controlled MFC also revealed Quick Anodophilic Activity Onset from the same Activated Sludge Inoculum

41

2.3.2.1 Evolution of Anodic Current after a short Lag-phase of only One Day 42

2.3.2.2 Anodophilic Bacteria: Biofilm instead of Suspended Cells 43

2.3.2.3 The Anodophilic Biofilm Established in the Potentiostatically Controlled MFC also gave similar Power Output as the Ferricyanide-Cathode MFC

43

2.3.2.4 SEM reveals only Thin and Low Biomass Density at the Highly Active Biofilm-Anode

45

2.3.3 Performance of a Dissolved Oxygen-based Cathode 46

2.3.3.1 Acidified Catholyte increases Open Circuit Voltage of the MFC 47

2.3.3.2 Phosphate Buffer improves the Dissolved Oxygen-Cathode Performance 48

2.3.3.3 Ferricyanide-Cathode Outperforms Dissolved Oxygen-Cathode 50

2.4 Conclusion and Implication 51

Chapter 3 Affinity of Microbial Fuel Cell Biofilm for the Anodic Potential 53

3.1 Introduction 54


3.2.1 Microbial Fuel Cell Start-up and On-line Process Monitoring 56

3.2.2 Calculation and Analysis 58

3.2.2.1 Determination of Voltage, Current and Power Generation 58

3.2.2.2 Measuring the Effect of Anodic Potential on Microbial Activity by varying External Resistance

58


3.3.1 Characteristics of the MFC 59

3.3.2 Initial Changes of Resistors result in Steady States of Different Microbial Activities

59

3.3.3 At a Certain Potential Range the Dependency of Microbial Activity on Anodic Potential is Less Defined

60

3.3.4 Apparent Maximum of Microbial Activity at Relatively Low Anodic Potentials 62

3.3.5 An Activity Maximum is only obtained when moving from Low to Higher Anodic Potentials

63

3.3.6 Open Circuit Drop in Anodic Potential 64

3.3.7 Detailed Interpretation of the Dependency of Microbial Activity on Anodic Potential

65

3.3.8 Redox Capacitance as an Explanation of Apparent Current Maximum 68

3.3.9 Implication of Findings on the Design and Operation of MFCs 68

Chapter 4 A New Approach for in-situ Cyclic Voltammetry of a Microbial Fuel Cell Biofilm without using a Potentiostat

69

4.1 Introduction 70


4.2.1 Underlying Principle of the Proposed Method 72

4.2.2 Microbial Fuel Cell 72

4.2.3 Cyclic Voltammetry by Feedback-Controlling the External Resistance 73

4.2.4 Cyclic Voltammetry using a Three-electrode Potentiostat 76


4.3.1 Feedback Controlling the External Resistance of a MFC enables Cyclic Voltammetry of MFC Biofilm

77

4.3.2 Limitations and Implications of the New Method 80

Chapter 5 An Anodophilic Biofilm prefers a Low Electrode Potential for Optimal Anodic Electron Transfer: A Voltammetric Study

82

ix

5.1 Introduction 83


5.2.1 Anodophilic Biofilm and Growth Medium 85

5.2.2 Construction and Operation of Bio-electrochemical Cell 85


5.3.1 Only Actively Metabolizing Biofilm Exhibited Electrochemical Activity in Cyclic Voltammetry

88

5.3.2 Redox Mediators of the Active MFC Biofilm were Located Far Away from the Electrode Surface

90

5.3.3 Detail Interpretation of the Anodophilic Properties of the MFC Biofilm in CV 91

5.3.4 Step-Change of Electrode Potential Signified the Existence of an Optimal Electrode Potential for Current Production from the Biofilm

93

5.3.5 Implication of Findings 95

Chapter 6 Alternating Bio-Catalysis of Anodic and Cathodic Reactions alleviates pH Limitation in a BES

97

6.1 Introduction 98


6.2.1 Electrochemically Active Biofilm and Growth Medium 101

6.2.2 Construction and Monitoring of the Bioelectrochemical System 101

6.2.3 Operation of the Bioelectrochemical Process 103

6.2.3.1 General Operation 103

6.2.3.2 Alternating Supply of Acetate and Oxygen to the Anodophilic Biofilm 103

6.2.3.3 Cathodic Electron Balances 105

6.2.4 Catalytic Effect of the Anodophilic Biofilm on Cathodic Oxygen Reduction 105


6.3.1 Acidification of Electrolyte diminishes Anodic Current Production 106

6.3.2 Alternating Supply of Acetate and Oxygen to a Biofilm limits Anodic Acidification

108

6.3.3 Operating the Described MFC at +200 mV did not enable a Cathodic Electron Flow

110

6.3.4 Cathodic Electron Balance 110

6.3.5 Catalytic Effect of the Anodophilic Biofilm on the Cathodic Reaction 111

6.3.6 Expected Power Production of MFC with Anodophilic Bacteria at both the Anode and Cathode

113

6.3.7 Implication of the Findings 113

6.3.7.1 Potential Benefits of the Proposed Concept to developing Sustainable MFC Processes

113

6.3.7.2 Anodophiles Catalyzing the Cathodic Reaction 114

Chapter 7 A Scalable Bioelectrochemical System for Energy Recovery from Wastewater — Rotatable Bio-Electrochemical Contactor (RBEC)

116

7.1 Introduction 117


7.2.1 Construction of the RBEC Reactor 118

7.2.2 Design of Two Half Discs Serving as Anode and Cathode 119

7.2.3 Process Monitoring and Control 121

7.2.4 Bacterial Inoculum and Synthetic Wastewater 123

7.2.5 Reactor Operation 123

7.2.5.1 Start-up as a Microbial Fuel Cell 123

7.2.5.2 Coupling the RBEC with a Power Source 124

7.2.6 Scanning Electron Microscopy of Biofilm-Electrode 124


7.3.1 Operation as a Microbial Fuel Cell for Electricity Generation 126

7.3.1.1 Increased Anodophilic Activity increases Power Output over Time 126

x

7.3.2 In Situ Supply of Oxidizing Power via Conductive Contactor (as Electron Flow) increases COD Removal

128

7.3.2.1 Parasitic Current decreases Coulombic Recovery 130

7.3.3 Sequential Flipping the Electrode Discs allows Alternate Current Generation

132

7.3.4 Coupling the RBEC with an External Power Source to achieve Higher Current

134

7.3.4.1 Electrochemically Assisted Anode Facilitates the Establishment of Anodophilic Biofilm

134

7.3.4.2 Sequential Flipping the Electrochemically-Assisted Discs establishes Anodophilic Biofilm on both Half Discs

136

7.3.5 Intermittent Flipping of the Discs avoids the Continuous Alkalization of the Cathode

137

7.3.6 The Established Biofilm could catalyze a Cathodic Oxygen Reduction 138

7.3.7 Energy Evaluation of the Electrochemically Assisted RBEC process 140

7.4 Concluding Remarks 140

Chapter 8 A Rotatable Bioelectrochemical Contactor (RBEC) enables Electrochemically Driven Methanogenesis

142

8.1 Introduction 143


8.2.1 Bacterial Seeding Inoculum and Synthetic wastewater 144

8.2.2 Anoxic Operation of the Reactor Headspace of the Electrochemically-Assisted RBEC

144

8.2.3 Gas Production and Measurements 146

8.2.4 Examination of Electrode Biofilm using Scanning Electron Microscopy 146


8.3.1 Anoxic Cathode enables Hydrogen Gas Formation 148

8.3.2 Methane instead of Hydrogen was the Predominant By-product after Prolonged Anoxic Operation

151

8.3.3 Is Hydrogen a Key End-product of the Cathodic Reduction Reaction in the Methane-Producing RBEC?

154

8.3.4 Energy Balance 156

8.4 Implication of Findings 157

Chapter 9 Conclusions and Outlook 158 9.1 Conceptual Progression of the Thesis 159

9.1.1 A Highly Anodophilic Biofilm can be Easily Established from Activated Sludge

159

9.1.1.1 MFC Power Output is Undermined by Limitations other than the Metabolic Capacity of the Bioanode

159

9.1.1.2 Wastewater Treating-BES requires Innovative R&D Approaches 161

9.1.2 Anodophilic Biofilm Affinity for the Anodic Potential 163

9.1.3 The Use of Computer Feedback for Dynamic BES Process Control 163

9.1.4 The Rotatable Bioelectrochemical Contactor: A New Option for Large-Scale Wastewater Treatment

164

9.1.4.1 RBEC as a MEC for Enhanced Cathodic Oxygen Reduction, Hydrogen or Methane Production

167

9.2 Final Remarks 168

References 169

Appendix 183 Author’s Curricular Vitae 198

- 1 -

1 Introduction and Aim

Chapter Summary

Every day, we turn ‗clean‘ water into ‗waste‘ water. For sanitation reasons, wastewaters

need to be treated in order to protect human health and the environment from water borne

diseases, and pollution such as eutrophication. However, traditional wastewater treatments such

as activated sludge processes are energy intensive.

Owing to the ever increasing concerns of global energy crisis and its associated

environmental impact, there is a growing need to develop a more energy-efficient wastewater

treatment process. A more radical goal is to recover useful energy from the treatment process.

Recently, bioelectrochemical systems (BESs): microbial fuel cell (MFC) and microbial

electrolysis cell (MEC) are well perceived as the prime technologies to achieve this goal (Logan

2005a; Lovley 2008; Rabaey and Verstraete 2005; Rozendal et al. 2008a).

In the past few years, significant research progress has already been made towards this

goal. However, as BESs are highly complex systems involving multi-disciplinary knowledge

such as microbiology, electrochemistry, material science and engineering, many hurdles still

need to be overcome before the technology becomes practical. The overall aim of this thesis is to

generate understanding that will be helpful in the development of an energy recovering

wastewater treatment process using BESs. The thesis is broadly divided into two main themes.

Firstly is the study of the fundamental aspect of BESs aiming to explore the potential of BESs

for electricity generation and to identify process bottlenecks. The second theme is to develop

practical solutions and to design a practical reactor configuration to overcome the established

bottlenecks.

This chapter aims at reviewing the current state of the art in research and development

related to energy recovery from wastewater using BESs. The basic principles of BESs are

outlined. Emphasis is given on BES that can convert organics into electrical energy directly (i.e.

MFC). Finally, some of the major obstacles of the technology are highlighted. The scope of the

thesis is presented in Section 1.2.

Chapter 1: Introduction and Thesis Aim

- 2 -

1.1 Background – Literature Review

1.1.1 We need to reduce Global Energy Consumption

Of the 13 terrawatts (1012

watts) energy consumed worldwide, approximately 80% (~10

TW) is derived from fossil fuels (e.g. crude oil and natural gas) (Johansson and Goldemberg

2004). Because their formation requires many millions of years, fossil fuels are considered as

non-renewable energy source. Combustion of fossil fuels leads to a continuous increase of

carbon dioxide level in the atmosphere, causing the well-known environmental effects of ―global

warming‖ and climate change (Chamberlain et al. 1982, pp.255; IPCC 2007; Klass 1993)

(Figure 1.1). These effects resulted in a rising sea level due to both expansion of sea water and

the melting of glaciers (IPCC 2007).

Figure 1.1 Historical and projected trends of atmospheric CO2 concentration and average air

temperature. Adapted from Rittmann (2008); original source: Intergovernmental Panel

on Climate Change, IPCC (2007).

Although in the past decade many countries have proposed various approaches to tackle

the problems, still the total world consumption of energy is recently projected to increase by

50% from 2005 to 2030 (EIA 2008). Energy demand in most developed nations (Organization

250

350

450

550

650

750

850

1800 1900 2000 2100Year

Ca

rbo

n D

ioxid

e

(pp

m v

ol.)

15

16

17

18

19

20

Te

mp

era

ture

( oC)

Atmospheric carbon dioxide

Average temperature

Year 2006


- 3 -

for Economic Co-operation and Development (OECD) countries) is predicted to expand at an

average annual rate of 0.7%, while energy consumption in the emerging economies of non-

OECD countries (e.g. China and India) is expected to expand at an even higher annual rate of

2.5% (EIA 2008). Of the expanded energy demand, 80-85% will still be derived from fossil

fuels (IEA 2004).

In contrast to the expanding global energy demand, the global supply of fossil fuels has

already peaked and is now declining. A theoretical study has estimated that crude oil and natural

gas reserves will be depleted by the year of 2070 (Klass 2003). Taking also the rapidly growing

world population and the increase of prosperity in developing nations (especially China and

India) into account, the global consumption and demand of energy will increase dramatically

before depletion of fossil fuels (Boeriu et al. 2005).

Clearly, the supply of fossil fuels is unlikely matched with our ever increasing energy

demand. The consequences will not only be an increased energy price leading to instability of

the global economy, but more seriously is the irreversible damage of our natural environment.

Hence, reducing fossil energy consumption and developing renewable alternatives are the prime

objectives of mankind as well as of this PhD thesis.

1.1.2 Wastewater Treatment: From ―Energy-to-Waste‖ to ―Energy-from-

Waste‖?

Wastewater treatment is of utmost importance to safeguard both human and ecological

health. In most developed and urbanized societies, approximately 1.5% of the total electrical

energy consumption is directed to their municipal wastewater treatment processes (Logan 2008).

Therefore, developing a more energy-efficient wastewater treatment can alleviate such energy

demand. In fact, most of the wastewaters (e.g. human, animal, and industrial) already contain a

large amount of chemical energy stored in the form of dissolved organic matter (Angenent et al.

2004). This energy not only can be harnessed for ―cleaning up‖ the wastewater itself, but it can

also be converted into other useful energy forms, available for human consumption again. As

dissolved organic matters in wastewaters are primarily derived from solar energy (via

photosynthesis), wastewaters can be seen as a potential renewable energy resource.


- 4 -

1.1.2.1 Activated Sludge Processes: Energy-to-Wastewater

Activated sludge (AS) process is currently the most widely adopted biological

(secondary) process for municipal wastewater treatment. However, it is an energy intensive

process in which air (oxygen) is actively transported (either by surface turbines or submerged

diffusers) from the gas-phase (atmosphere) into the liquid-phase (wastewater) in order to

maintain the wastewater aerobic and the activated sludge (a highly concentrated bacterial

consortium) in suspension. The activated sludge degrades the organic contaminants in the

wastewater by using dissolved oxygen as their terminal electron acceptor. For each m3 of

wastewater approximately 8 m3 of air is required (Lee and Lin 2000). Such active aeration

process consumes tremendous amount of energy (about 1 kWh per kg COD removed or about

0.7 kWh m-3

), which commonly leads to about 60-80% of the total operational cost of the entire

wastewater treatment process. Therefore, in order to reduce treatment costs many studies have

been investigating ways to optimize aeration efficiency or to minimize aeration without

compromising treatment efficiency (Chachuat et al. 2005; Chambers et al. 1998).

1.1.2.1.1 Bringing Oxygen from Air to Wastewater requires Energy

The energy demand of aeration in AS processes lies fundamentally on the mass transfer

of air to water. A certain amount of mechanical energy must be supplied to transport a unit

volume of air into a unit mass of wastewater. This energy is directly proportional to the height of

the aeration basin, and hence the depth of the wastewater (Figure 1.2). However, due to limited

oxygen transfer efficiency (OTE) only part of the oxygen in the transferred air is accessible to

the bacteria in the wastewater. Although OTE can be maximized by reducing the size of the air

bubble (Table 1.1), it appears that at least 70% (assuming a maximal OTE of 30% according to

Table 1.1) of the oxidizing power (as oxygen) is inevitably lost in the process. Such inevitable

loss represents a waste of energy in the treatment process.

1.1.2.1.2 Electron Transfer instead of Mass Transfer

Instead of transferring oxygen from a gas phase into a liquid-phase where the bacteria

and electron donors (i.e. organics) are located, ―oxygen-equivalents‖ required for the organic

removal in a wastewater can be indirectly supplied in the form of an electron-accepting (i.e.

oxidizing) surface such as an inert electric conductive electrode in BES processes.


- 5 -

Figure 1.2 Aeration energy is directly proportional to water depth. Note: assumes standard

oxygen transfer rate 60 kg O2 h-1

, air blower efficiency 80%, wastewater basin volume

500 m3, air flow rate 1 m

3s

-1, chemical oxygen demand (COD) of wastewater 320 mg L

-1.

Table 1.1 Standard oxygen transfer efficiency (OTE) and oxygenation efficiency (OE) in

activated sludge processes

Bubble Size Bubble Diameter Standard OTE OE

mm % kg O2/kWh

Fine <3 10 to 30 1.2 to 2

Medium 3 to 6 6 to 15 1 to 1.6

Coarse >6 4 to 8 0.6 to 1.2

Adapted from (von Sperling and de Lemos Chernicharo 2005, page 479)

Since electrons are virtually weightless, the transfer of electrons via an electrode does not

incur any mass transfer energy as in AS aeration processes. Further, the fact that oxygen is not

required for the direct oxidation of organics as well as the possibility of direct recovery of

electrical energy from a wastewater treatment process are features that may increase the overall

energy efficiency of a wastewater treatment process. Therefore, BESs are emerging as an

innovative option for wastewater-to-energy in the future.

0.0

0.5

1.0

1.5

2.0

0 2 4 6 8 10

Ae

ratio

n E

ne

rgy

(kW

h· k

g C

OD

-1)

Water Depth (m)


- 6 -

1.1.2.2 Options for Wastewater-to-Energy

Prior to BESs, two other biological processes have been proposed for the conversion of

chemical energy in a wastewater to other useful forms of chemical energy (e.g. biogas (methane

+ CO2), hydrogen gas); (1) methanogenic anaerobic digestion and (2) biological fermentative

hydrogen gas production (also known as dark hydrogen fermentation). The by-products of these

two bioprocesses (i.e. biogas/ hydrogen) are eventually converted into electrical energy using

heat combustion engines or hydrogen fuel cells, respectively.

1.1.2.2.1 Methanogenic Anaerobic Digestion

Methanogenic anaerobic digestion is an established technology for treating a variety of

organic wastes including domestic wastewaters (de Mes et al. 2003). Currently, a total of 2,266

registered full scale high-rate methanogenic digesters are currently operating worldwide (van

Lier et al. 2008). Compared to aerobic wastewater treatment, anaerobic treatments generally

consume less energy. Depending on the need for pumping and recycling effluent, the energy

requirements at ambient temperature are in the range 0.05-0.1 kWh m-3

(0.18-0.36 MJ m-3

) (~0.7

kWh m-3

for AS process) (de Mes et al. 2003).

In fact, the generation of biogas enables net energy recovery from the process. For each

kg COD removed about 13.5 MJ CH4 energy is produced, giving 1.5 kWh high-grade electrical

energy (assuming 40% electric conversion efficiency) (van Lier et al. 2008).

Apart from the possibility of recovering useful energy (as biogas) from a wastewater

instead of investing energy in the process, anaerobic wastewater treatment has an additional

advantage over aerobic wastewater treatment — low sludge production. About 5% of the COD

is converted into sludge in anaerobic processes compared to about 30-60% in aerobic processes

(Figure 1.3). Less sludge handling can reduce the overall treatment cost. Further, the residual

anaerobic sludge even has a positive market value as they can be used as the seeding material to

start up another methanogenic treatment process (van Lier et al. 2008).

Nevertheless, the effluents from anaerobic digester typically still contain a high level of

organics (ranging from 0.5 to a few grams of residual volatile fatty acids) (van Lier et al. 2001).

These dead-end organic compounds represent a pool of unrecovered chemical energy in

anaerobic digestion processes. BESs are perceived as the suitable processes to deal with such


- 7 -

low organic waste stream (Clauwaert and Verstraete 2008; Rozendal et al. 2008a), and hence,

rather than competing with the existing anaerobic digestion processes, BESs can be seen as a

complementary technology to anaerobic digestion (Pham et al. 2006).

Figure 1.3 Carbon and energy balance in aerobic and anaerobic wastewater treatments.

(Adapted from van Lier et al. (2008))

1.1.2.2.2 Fermentative Hydrogen Production

Fermentative hydrogen production (also known as dark fermentation), on the other hand,

is an emerging waste-to-energy process mainly targeted towards the concept of ―hydrogen

economy‖, a hypothetical future economy in which the primary form of stored energy is

hydrogen (Momirlan and Veziroglu 2005; Rifkin 2002). The underlying principle of

fermentative hydrogen production is very much alike methanogenic anaerobic processes except

that the former process requires inhibition of hydrogen-consuming microorganisms

(hydrogenotrophic methanogens). However, due to thermodynamic constraint of the involved

biochemical pathways, only 10 to 20% of the stoichiometric maximal hydrogen yield (i.e.

maximal 12 moles of H2 per mole of hexose) could be achieved in practice. I have carried out a

review on this topic which has been published (Cheng et al. 2007, see Appendix 1). A more

Aerobic Wastewater

Treatment

Aeration Energy

100 kWh

Influent

100 kg COD

Heat Loss

Sludge

30-60 kg COD

Effluent

10-12 kg COD

Anaerobic Wastewater

Treatment

Influent

100 kg COD

Biogas 40-45 m3 (~70% CH4)

Sludge

5 kg COD

Effluent

10-20 kg COD


- 8 -

comprehensive review of fermentative hydrogen production as a wastewater-to-energy method

can also be found in other reviews (Adams and Stiefel 1998; Benemann 1996; Hallenbeck and

Benemann 2002; Hallenbeck and Ghosh 2009).

Although there is a significant industrial demand for hydrogen, producing hydrogen from

wastewater treatment via fermentation with an aim to recover renewable energy is not

appropriate. Compared to methane, hydrogen has a much lower density and boiling point, only

0.09 kg Nm-3

and -252.9 °C, respectively (vs. 0.717 kg Nm-3

and -182.5 °C, respectively for

methane). These physical properties make hydrogen gas less suitable for transport and storage

compared to methane.

It appears that the popular equation ―H2 + O2 → electricity + H2O‖ has somewhat

oversimplified the ―hydrogen economy‖ concept (Bossel 2006). In reality, the equation should

be modified as ―[H2]initial

+ energies… → [H2]final

+ O2 → electricity + H2O‖. From a point

where hydrogen gas is produced (i.e. [H2]initial

) to a point it is converted into electricity (i.e.

[H2]final

), many energy-consuming intermediate steps are involved. These extra energies can only

be obtained from other energy sources (either fossil fuels or other renewable), but certainly not

from the primary hydrogen production step. In other words, a biologically produced hydrogen-

rich biogas requires energy intensive post-treatment such as purification and pressurization

(from 1 bar to about 600 bars) before it is finally delivered to a fuel cell for electricity generation.

How do we justify the amount of CO2 released if these extra energies have to be derived from

nonrenewable fossil fuels?

Hydrogen generation using bioelectrochemical system (BES) has been proposed in the

literature since 2005 (Liu et al. 2005; Rozendal et al. 2006b). In these pioneering studies, (non-

fermentable) acetate was successfully used for the generation of hydrogen leading to an overall

increase in hydrogen yield. BES is therefore considered as a very promising technology for

turning wastewaters into a useful form of energy (apart from hydrogen, even electricity can be

directly recovered from wastewater using BES), and it will be introduced in the remaining

sections of this chapter.


- 9 -

1.1.3 Bioelectrochemical Systems for Sustainable Wastewater-to-Energy

In nature, energetic of biological systems involves electron flow. Electrons flow from an

electron donor that becomes oxidized, to an electron acceptor that becomes reduced. Hence, it is

possible to manipulate biological processes involved in a wastewater treatment process using

electrochemical methods.

The early discoveries of a bacterial capacity to oxidize dissolved organic matters using

an electron accepting electrode (Davis and Yarbrough 1962; Potter 1911) had provoked the

recent development of bioelectrochemical systems (BESs), including microbial fuel cell (MFC)

and microbial electrolysis cell (MEC) processes. BESs allow a direct conversion of chemical

energy stored in a chemical compound into a useful form of energy, with MFC used for a net

production of electricity and MEC for a production of hydrogen gas as a by-product (Rozendal

et al. 2008a).

Recovering energy from wastewaters using BESs is promising because of their high

energy conversion efficiencies compared to all other energy conversion bioprocesses (Rabaey et

al. 2005c). Further, BESs can treat wastewaters that are not suitable for anaerobic digestion

processes (e.g. low-strength wastewater, wastewater composed with mainly volatile fatty acids,

etc) (Rittmann 2008; Watanabe 2008).

Similar to anaerobic digestion, BES also enables removal of organics (anodic oxidation)

with a much lower sludge generation yield (ranging from 2.4 to 26.5 times lower) compared to

aerobic activated sludge processes (Clauwaert et al. 2008a). For instance, a yield of 0.02-0.05 g

biomass-C (g-substrate-C)-1

with acetate as the electron donor substrate (Aelterman et al. 2008);

0.07-0.22 g biomass-COD (g substrate-COD)-1

with glucose as the electron donor substrate

(Rabaey et al. 2003); while 0.53 g biomass-COD (g substrate-COD)-1

for activated sludge

wastewater treatment processes (Verstraete and van Vaerenbergh 1986). Hence, BESs are

emerging as a clean technology for energy recovery from wastewaters.


- 10 -

1.1.4 Overview of Bioelectrochemical Systems

1.1.4.1 BES: A Fast Growing Field of Research

In just a few years course of the author‘s PhD study (2006 - 2009), BES research has

been rapidly expanding in the scientific community worldwide. This results in a significant

increase in both the number of published reports and our current knowledge of BES, making it a

challenging task to prepare an up-to-date literature review on the topic. Hence, readers are

kindly referred to the current literature for a more comprehensive overview of specific topic of

BES. In particular, several excellent and extensively cited articles are recommended: (Clauwaert

et al. 2008a; He and Angenent 2006; Logan et al. 2008; Logan et al. 2006; Lovley 2006b; Pham

et al. 2006; Rabaey et al. 2005c; Rabaey and Verstraete 2005; Rozendal et al. 2008a; Schroder

2007).

The following sections review the underlying principles of BES that are prerequisites for

this thesis, and pinpoint the major limitations hindering the up-scaling potential of the BES.

Only BESs using bacteria as the electrochemical ―living catalyst‖ are concerned here.

1.1.4.2 Basic Features of BES

To realize the conversion of chemical energy from wastewater into electrical energy

(electron flow), all BES should have several basic features (Figure 1.4). Firstly, an electron

donor (e.g. organics present in a wastewater) is oxidized by an electrochemically active

anodophilic biofilm which can subsequently transfer the liberated electrons to an electrode (here

anode). Secondly, the electrons at the anode are driven by a potential gradient toward a counter

electrode (here cathode) via an external conductive circuit. In MFC, such electron flow

generates a net electrical energy if a suitable resistive load such as a resistor is located in the

circuit. The MFC cathode can spontaneously reduce an oxidized species (e.g. oxygen). While in

MEC, an external electrical energy from a power source (e.g. DC supplies or potentiostats) is

required to trigger a cathodic reduction (e.g. reduction of protons for H2 formation).

In BES the cathodic reduction can be of either abiotic or biotic nature, where a suitable

bacterial culture (usually exists as a biofilm) catalyzes the electron flow from the cathode to a

terminal electron acceptor (Clauwaert et al. 2007b; Rabaey et al. 2008; Rozendal et al. 2008b).

To date, the so-called biocathode has become a preferred option for BES operation as the need


- 11 -

of using expensive and non-sustainable metal-based chemical catalysts (e.g. platinum) can thus

be excluded (He and Angenent 2006; Rabaey et al. 2007).

Figure 1.4 Schematic drawing of the basic features of microbial fuel cell (MFC) and microbial

electrolysis cell (MEC). Red.= reduced bacterial substrate (electron donor); Ox.=

oxidized electron acceptor.

Finally, in order to sustain electron flow (flow of negative charge) from the anode to

cathode in the BES the ionic charges between the two half cells must be balanced (neutral). In

other words, for each electron flowing across the external circuit, either a positively or a

negatively charged ionic species must be transferred from the anode to the cathode or vice versa

(Figure 1.4).

It is worth to point out that in most reported BES studies so far, ion exchange membranes

(e.g. cation exchange membranes, CEMs; anion exchange membranes, AEMs; bipolar

membranes, BPMs) are used to physically separate the reductants (at anode) and oxidants (at

Direction of electron flow

Cation

Red. Ox.Ano-

Bact.

Cath-

Bact.

(Optional)

Anode

Anionor

Cathode

e- e-

e- e-

e- e-

Ion exchange membrane(Optional)

MFC: Power User

(e.g. Resistor)

MEC: Power Supply

Electrolyte Electrolyte


- 12 -

cathode) and to selectively allow ion mobilization between the two half cells. However, it can

also be excluded. Some recent studies have proven that both MFC and MEC could operate

without an ion exchange membrane (Call and Logan 2008; Clauwaert and Verstraete 2008; Hu

et al. 2008; You et al. 2007).

Indeed, apart from a high capital cost the use of membranes has several drawbacks on

BES‘s performance. The most well known limitations are the increase in ohmic resistance and

the establishment of a membrane pH gradient (Harnisch et al. 2008; Rozendal et al. 2006a). This

issue will be further discussed in Section 1.1.5.1.

Now, we can move on to the fundamental principle of how and why electrons can flow

in BES.

1.1.4.3 Principles and Thermodynamics of Bioelectrochemical Conversion

in BESs

The likelihood of any chemical reaction to occur in an electrochemical system under

standard or actual conditions can be evaluated by calculating the Gibb‘s free energy change (ΔG)

of the reaction (Bard et al. 1985; Chen 2003). Similarly, this principle can also be applied in

BESs. Electrical energy in BES is spontaneously generated only when the overall reaction is

thermodynamically (or energetically) favorable (ΔG < 0). Yet, it may be more convenient to

evaluate the reaction in terms of the overall cell electromotive force, Eemf (V), which is defined

as the potential difference between cathode (i.e. positive terminal) and anode (i.e. negative

terminal) of the BES:

Eemf (V) = E cathodic (V) – E anodic (V) (eq. 1-1)

Where, E cathode = emf of a specific reaction taking place at cathode

E anode = emf of a specific reaction taking place at anode

The maximum work (Wmax) that a system can perform is related to the Gibbs free energy

changes (∆G). This can be estimated from Eemf according to the following relationship (Logan et

al. 2006):


- 13 -

Wmax = Eemf·Q = Eemf·n·F= -∆G (eq. 1-2)

Where, Wmax = maximum theoretical work

Eemf = potential difference between the anode and cathode

Q = charge transferred in the reaction, expressed in Coulomb (C)

n = number of electrons per reaction

F = Faraday‘s constant (96,485 C mol-1

)

Rearranging equation 1-2 gives:

Eemf = − ∆G

nF (eq. 1-3)

Under standard condition (i.e. 298K, 1 atm pressure, 1 M concentration for all species) gives:

Eemf𝑜 = −

∆G0

nF (eq. 1-4)

Where E0

emf = standard cell electromotive force. Based on the above equations, the overall

electromotive force of a particular reaction under any specific condition can be calculated

according to the Nernst equation (eq. 1-5):

Eemf = Eemf𝑜 −

RT

nFln(П) (eq. 1-5)

Where, R = universal gas constant (8.314 x 10-3

kJ∙mol-1

∙K-1

);

T = absolute temperature (K);

П = reaction quotient calculated as the activities (i.e. concentrations at dilute

concentrations) of the products divided by those of the reactants.

For a reaction of: aA + bB → cC + dD, the reaction quotient (П) can be defined as the

following (Bard and Faulkner 2001):

П =[𝐶]𝑐 [𝐷]𝑑

[𝐴]𝑎 [𝐵]𝑏 (eq. 1-6)


- 14 -

1.1.4.3.1 MFC and MEC: Similarities and Differences

1.1.4.3.1.1 Similarities: Anodic Oxidation is catalyzed by Microbes

The first step of the microbial transformation of chemical energy into electrical energy in

both MFC and MEC is similar (Figure 1.5). In which, a reduced (more negative redox potential)

electron donor substrate is taken up by the bacteria located at close proximity (or directly

attached) to an anode, which has a higher (more positive) redox potential compared to the

bacteria itself. The bacteria will then catabolize the electron donor via a series of biochemical

redox reactions inside the cell (e.g. TCA cycle). In these redox processes, the bacteria must

conserve a certain amount of energy (i.e. Gibbs free energy) to meet their cellular needs (e.g. for

cell maintenance and growth) (Thauer et al. 1977).

Figure 1.5 A schematic diagram showing the thermodynamics of microbial fuel cell (MFC) and

microbial electrolysis cell (MEC). Note: This representation neglects electrochemical

losses (also known as overpotential) based on ohmic resistances (electrode and

electrolyte), concentration (mass transfer limitation) and activation polarizations and

kinetic constraints.)

Redox Potential (V)

+ve

e- Donor

(Acetate)

Microbe

Anode

-0.28

+0.82

ΔGo’Elec= -ve

-0.42 e- Acceptor

(H+/ H2)

ΔGo’bacteria = -ve

e- Acceptor

(O2/ H20)

Energy

Input

Energy

Output

ΔGo’Elec= +ve

-ve

MEC

MFC

Load(e.g. resistor)

Power supply

Cathode

ΔGo'Max. Output = -ve

ΔGo'Min. Input = +ve

1.1 V

0.14 V

Anion

Cation

or


- 15 -

1.1.4.3.1.1.1 Dilemma: A Lower or a Higher Anodic Potential?

From our energy recovery standpoint, such bacterial energy gain (denoted as ΔGo‘

bacteria

in Figure 1.5) can be considered as an ―energy loss‖ because it cannot be transformed into

electrical energy in the BES (Schroder 2007). Thus, a dilemma exists when determining which

processes (bacterial energy gain? or an anode with a lower anodic potential?) should be favored

more when operating a BES (Figure 1.6).

Nevertheless, it is clear that in order to maximize electricity generation in MFC or to

minimize the extra energy input in MEC (both are denoted as ΔGo‘

Elec in Figure 1.5), a lower

amount of ΔGo‘

bacteria is preferred as it keeps the anode at a lower redox state and hence the

anode can be more reducing. Further, if the bacterial energy gain becomes too large (at ―too high‖

anodic potential), the electrical energy output may become diminished and electrons from the

substrate may be diverted toward an assimilation metabolism, leading to excessive biomass yield

but reduced electron recovery. Therefore, if the anodic bacteria could still benefit from a small

potential gradient (between the electron donor substrate and the anode) a lower operating anodic

potential is preferred.

Figure 1.6 Effects of anodic potential on the theoretical cell voltage and bacterial energy gain

(Gibbs free energy change, ΔGo‘

). Notes: ΔGo‘

data refer to standard condition, pH 7 and

298K; cell voltage (V) = cathodic potential – anodic potential.

-1.0

-0.5

0.0

0.5

1.0

1.5

-600 -400 -200 0 200 400 600

Anodic Potential (mV vs. Ag/AgCl)

ΔG

o'(k

J m

ol -1)

-900

-800

-700

-600

-500

-400

-300

-200

-100

0

100

Ce

ll V

olta

ge

(V

)

Cell Voltage: Coupled with an O2/H20 Cathode (+0.62 V |Ag/AgCl)

Gibbs free energy change

Cell Voltage: Coupled with a Poor Cathode (0 V |Ag/AgCl)

Ba

cte

rial E

ne

rgy G

ain

BE

S E

ne

rgy O

utp

ut


- 16 -

Apparently, the above consideration could lead to important research questions… e.g.

―How low (negative) can an operating anodic potential be maintained and yet still allow a

reasonably high (if not maximal) microbial activity (as current)?‖ ―Can we establish an anodic

microbial consortium that has a suitable metabolic and electron transfer pathway to effectively

utilize a very negative anodic potential?‖

1.1.4.3.1.2 Difference: Exergonic or Endothermic Cathodic Reduction?

As depicted in Figure 1.5, the fundamental difference between MFC and MEC lies in the

thermodynamic nature of their cathodic reaction. In MFC, the cathodic potential established by

the reducing power from the anode is sufficient to drive a cathodic reduction of a terminal

electron acceptor (e.g. oxygen, nitrate or ferricyanide) (reaction 1-2 to 1-4 in Table 1.2). Such a

downhill (exergonic) redox reaction (ΔGo‘

Max. Output < 0) enables a spontaneous energy output (as

electricity) if a suitable electrical load (e.g. resistor) is located in the external circuit of the MFC.

Table 1.2 Thermodynamic evaluation of BES with different cathodic reactions (assuming an

anodic acetate oxidation) under standard condition (pH 7, 298K).

Eo’ (V)

a

BES Eemf (V)

BES ΔGo’

(kJ mol-1

) Reaction

Anodic Oxidation

CH3COO- + H2O → CO2 + 7H

+ + 8e

- -0.28 --- --- (1-1)

Cathodic Reduction

O2 + 4H+ + 4e

- → 2H2O +0.82 +1.10 -849 (1-2)

NO3- + 6H

+ + 5e

- → 0.5N2 + 3H2O +0.74 +1.02 -787 (1-3)

Fe(CN)63-

+ e- → Fe(CN)6

4- +0.36 +0.64 -494 (1-4)

H+ + e

- → H2 -0.42 -0.14 +108 (1-5)

a Values adapted from He and Angenent (2006); Eemf values are calculated according to eq. 1-1, Eemf

= E cathodic – E anodic.

On the other hand, the cathodic reduction in MEC requires a more reducing (i.e. negative)

cathodic potential, which cannot be satisfied with the reducing power of the anode (e.g. reaction

1-5 in Table 1.2). An extra amount of reducing power (here a potential of -0.14 V for H2

formation) is required from an external power source (e.g. potentiostat) to kick off the cathodic


- 17 -

reaction (ΔGo‘

Min. input > 0). So far, MEC is mostly established for the cathodic generation of

hydrogen gas (i.e. cathodic reduction of protons).

As electrochemically active bacteria catalyze the release of electrons from an organic

electron donor at a higher energy level (i.e. at low (i.e. more negative) anodic potential), the

amount of energy required for hydrogen production is lowered compared to that in a

conventional hydrogen production using water electrolysis method. Typically, such conventional

process requires 1.8 to 2.0 V to electrolyze water. Hence, using MEC to produce hydrogen has

attracted remarkable attention from the industry (Rozendal 2007).

1.1.4.4 Microbial Oxidation of Organics Using an Insoluble Electrode as

Electron Acceptor

Arguably, the most essential phenomenon in BES systems is the capacity of bacteria to

transfer their electrons exocellularly to an insoluble electrode. This bacterial capacity is essential

as it is the key step of converting checmial energy into an electrical energy in all BES systems.

In nature, exocellular electron transfer can be observed in various geochemical environments.

The most well studied examples are the microbial reduction of insoluble Iron (III) oxide (e.g. by

genera of Geobacter) and Manganese (IV) oxide (e.g. by genera of Shewanella) commonly

occuring in anaerobic habitats such as groundwater aquifers and freshwater or marine sediments

(Lovley et al. 2004; Myers and Nealson 1990; Nealson and Saffarini 1994).

Such electron transfer enables the bacteria to conserve energy (e.g. via TCA cycle, ATP

generation by electron transport phosphorylation) to meet their metabolic need. The maximal

amount of energy that the bacteria can possibly conserve is in theory directly proportional to the

redox potential difference between the electron donor (reductant) and the electron acceptor

(oxidant) (Table 1.3).

Table 1.3 Standard reduction potentialsa of various redox couples that are of biological

interest Oxidation/Reduction Couples (mV vs. SHE) (mV vs. Ag/AgCl)

CO2/ Glucose -430 -630

H+/H2 -420 -620

NAD+/ NADH -320 -520

CO2/ Acetate -280 -480

S0/ HS

- -270 -470

SO42-

/ H2S -220 -420

Pyruvate2-

/ Lactate2-

-185 -385


- 18 -

2,6-AQDS/ 2,6-AHQDS -184 -384

Menaquinione oxidized/ reduced -75 -275

Pyocyanin oxidized/ reduced -34 -234

Methylene blue oxidized/ reduced +11 -189

Fumarate2-

/ Succinate2+

+31 -169

Thionine oxidized/ reduced +64 -136

Cytochrome b (Fe3+

)/ Cytochrome b (Fe2+

) +75 -125

Fe(III) EDTA/ Fe(II) EDTA +96 -104

Ubiquinone oxidized/ reduced +113 -87

Cytochrome c (Fe3+

)/ Cytochrome c (Fe2+

) +254 +54

O2/ H2O2 +275 +75

Fe(CN)63-

/ Fe(CN)64-

+430 +230

Fe(III) citrate/ Fe(II) citrate +372 +172

NO3-/ NO2

- +421 +221

NO3-/ N2 +740 +540

O2/H2O +820 +620 aStandard biological potential: pH7, 298K. Data mostly from Lehninger et al. (1993).

1.1.4.4.1 Exocellular Electron Transfer and Mediators

Exocellular electron transfer enables microbial oxidation of electron donors (e.g.

organics) using an insoluble electron acceptor such as an anode in BES. Bacteria capable of

catalyzing an electron transfer from a reduced electron donor to an anode are termed as

anodophilic bacteria (throughout this thesis). Other terminologies are also used in the literature.

For example, electricigen (Logan 2008); anode-respiring bacteria (ARB) (Torres et al. 2007);

electrode reducing microorganisms or electrode reducers (Lovley 2008)).

Although earlier studies have emphasized the important role of artificial redox mediators

(e.g. neutral red (Park et al. 2000; Park and Zeikus 2000); 9,10-anthraquinone-2,6-disulfonic

acid (AQDS), safranine O, methylene blue (Sund et al. 2007)) on exocellular electron transfers

in BES (mainly MFC), recent evidence has confirmed that bacteria (in either pure or mixed

culture studies) could establish their anodophilic capacity without using artificial mediators

(systems commonly known as mediator-less MFC) (Chaudhuri and Lovley 2003; Gil et al. 2003;

Kim et al. 2002; Rabaey et al. 2004).

For example, Kim and coworkers (2004) observed in a confocal scanning laser

microscopy study that the microorganisms in the biofilm attached on the anode of their MFC

were densely stained with florescent, indicating a majority of viable bacteria in the biofilm (Kim

et al. 2004a). Since no electron acceptors were supplied except the anode, the survival of those

microorganisms that were not in direct contact with the electrode implies that the bacteria have

developed alternative means to conserve energy using the anode without using artificial

mediators.


- 19 -

In fact, from a wastewater treatment standpoint the use of artificial mediators is

impractical as they need to be regularly replenished and may cause water contamination. Hence,

this thesis only considers self-mediated exocellular electron transfer.

1.1.4.4.1.1 Self-Mediated Exocellular Electron Transfer

In order for the anodophilic bacteria to self-mediate electron transfer to an anode, a

―redox link‖ must be developed between the bacteria and the anode. According to Schroder

(2007), an effective redox species must have several basic criteria:

(i) It must be able to physically interact with the electrode surface;

(ii) It must be electrochemically active and exhibits low oxidation overpotential at

the electrode surface;

(iii) Its standard potential should be as close to the redox potential of the electron

donor substrate of the bacterium as possible, or at least be significantly negative

to its final electron acceptor.

Figure 1.7 Exocellular electron transfer enables microbial oxidation of electron donors (e.g.

organics) using an insoluble electron acceptor such as an anode in BES. (A) redox active

mediator(s) serve as electron shuttle; (B) direct contact between bacteria and insoluble

electron acceptor without using soluble mediator(s).

Fe3+ + e- → Fe2+ or;

Mn4+ + 2e- → Mn2+ or;

Anode (oxidized) → Anode (reduced)

Medox

Medred

Bacterium

(1)

(2)

(3)(4)

(5)E donor

Medox

Medred

(7)

(2)

(3)(4)

(5)

(6)

Bacterium

(1)E donor

(1)

(1)

Inso

lub

le e

-Acce

pto

r

(Fe

3+; M

n4

+o

r An

od

e)

e-

A. ExocellularMediator Dependent

(Endogenous or exogenous):

B. ExocellularMediator Independent

(Mediatorless):

Electron flow in BESs

Examples:


- 20 -

Remarks: (1) Bacterium acquires electrons from electron donors (i.e. dissolved organic matters);

(2) Bacterium transfers the electrons to an oxidized mediator;

(3) Reduction of the oxidized mediator;

(4) Transfer of electrons from the reduced mediator to an insoluble electron acceptor;

(5) Regeneration of the oxidized mediator;

(6) Electron transfer from a reduced mediator produced by one bacterium to another

bacterium (Pham et al. 2007);

(7) Bacterium transfers electrons directly to an insoluble electron acceptor.

To our current knowledge, anodophilic bacteria may establish such ―link‖ with one of

the following two mechanisms (Figure 1.7): (1) direct physical contact with the anode using a

―link‖ located at the outer membrane (e.g. c-type cytochromes in Geobacteraceae, (Richter et al.

2009); or electrically conductive pili ―nanowires‖, which allow the bacteria located at a longer

distance away from the anode but still able to directly transfer electrons to it (Reguera et al.

2005)); and (2) self-secretion of a soluble redox active shuttle (mediator) that can mediate

electron transfer through diffusion and/or electrostatic interaction (Rabaey et al. 2007).

1.1.5 Limitations in BES Processes

Ongoing BES research efforts mainly focus on fundamental understanding of how

bacteria interact with an insoluble electrode and developing various configurations for different

application. (Clauwaert et al. 2008a; He and Angenent 2006; Logan and Regan 2006b; Logan et

al. 2008; Logan et al. 2006; Lovley 2006b; Rabaey et al. 2007; Rabaey and Verstraete 2005;

Rismani-Yazdi et al. 2008; Rittmann 2008; Rozendal et al. 2008a; Schroder 2007).

In particular, two fundamental limitations are emphasized: (1) the poor cathodic reaction;

and (2) the build-up of a pH gradient between an anode and a cathode over time in non-pH

buffered or poorly pH buffered systems. This so-called pH splitting phenomenon, if not

alleviated, can almost entirely block the electron flow and operation of the MFC.

1.1.5.1 The Use of Membrane Separators leads to pH Splitting Phenomenon

Ideally, either a proton (H+) or a hydroxide ion (OH

-) serves as the migrating ionic

species across an ion exchange membrane (refer to Figure 1.4). However, it becomes well

accepted in the literature that under operating conditions (i.e. pH close to neutral) prevailing in


- 21 -

most MFC systems, and also in MEC where protons are continuously consumed for the

generation of hydrogen gas at a cathode, the concentrations of both H+ and OH

- ions are several

order of magnitudes lower compared to most other ionic species in the electrolyte (Harnisch et al.

2008; Rozendal et al. 2008a). This results in a preferential migration of other ionic species to

maintain electroneutrality of the system, causing acidification and alkalization at the anode and

cathode, respectively. At present, there is no real solution available in the literature to

sustainably overcome such membrane associated pH splitting limitation yet (Harnisch and

Schroder 2009).

1.1.5.1.1 pH Splitting reduces BES Performance

pH splitting phenomena could lead to as high as a 6 pH units difference between the two

electrodes (Clauwaert et al. 2008b), severely decreasing the driving force of the system (i.e.

potential difference between anode and cathode) as each unit of pH gradient represents an

overpotential (i.e. potential loss) of 59 mV according to Nernst equation Consequently, the

current is diminished and hence diminishing the wastewater treatment performance.

1.1.5.1.1.1 A Fundamental Understanding on the Effect of pH Change on BES

Performance

The cathodic oxygen reduction in MFC and the cathodic hydrogen formation in MEC are

both proton consuming reactions, i.e.

O2 + 4H+ + 4e

- → 2H2O

H+ + e

- → H2

To understand the effect of pH on the BES performance, the reduction potential of these

cathodic reactions can be modeled according to Nernst equation (eq. 1-5). Since the H+ and OH

-

are always in equilibrium with each other through the water dissociation (Kw=[H+][OH

-] ≈ 10

-14),

net consumption of proton represents a hydroxide ion production. Hence, it is expected that the

localized pH at the cathode disc would increase over time. Assuming the concentration of other

reacting species (e.g O2 and H2) in the above cathodic reactions stay constant over time, then the


- 22 -

reduction potentials are solely dependent on the [OH-] and hence the corresponding Nernst

equation can be described as e.q. 1-7 (Rozendal 2007):

Ecathodet>0 = Ecathode

t=0 − RT

nFln

[OH −]t>0

[OH −]t=0

a (e.q. 1-7)

Where Ecathodet=0 and Ecathode

t>0 are the cathodic potential at time = 0 and time > 0, respectively.

The ―a‖ in e.q. 1-7 is the reaction coordinate of the hydroxyl ions in the cathodic reaction. Since

the formation of OH- is directly proportional to the current of the BES (i.e. 1 mol of OH

- is

produced per 1 mol of electron flow across the external circuit), the reaction coordinate, a, in e.q.

1-7 must be exactly equal to ―n‖. Hence, e.q. 1-7 can be further simplified as follow (e.q. 1-8):

Ecathodet>0 = Ecathode

t=0 − RT

Fln

[OH −]t>0

[OH −]t=0 (e.q. 1-8)

Based on this equation, for each unit of pH increase at the cathode (i.e. a 10-fold increase in OH-

concentration) after a certain period of time (t>0), the cathodic potential is decreased by 0.059 V

(i.e. 59 mV) (Figure 1.8):

Ecathodet>0 = Eca thode

t=0 − RT

Fln

[OH −]t>0

[OH −]t=0 (e.q. 1.9)

=Ecathodet=0 −

8.314 298

96485ln 10

= Ecathodet=0 − 0.059 V

As shown in Figure 1.8, alkalinization of the cathode decreases the cathodic potential. If

we assume the anodic reaction of a BES stay constant, a decrease of the cathodic potential

signifies an overall decrease of the driving force (e.m.f.) of the BES (refer to section 1.1.4.3 e.q.


- 23 -

1-1). As a consequence, this would diminish electrical energy output of a MFC, or a higher

energy input is required to bring the cathode to the lower potentials in order to trigger the

cathodic reduction in a MEC.

On the other hand, Figure 1.8 suggests a beneficial effect of pH change on the cathodic

potential when the cathode becomes more acidic instead of alkaline. An increase of +59 mV in

the cathode is obtained if the cathode was acidified by one pH unit. However, from a practical

viewpoint, as both the oxygen reduction and hydrogen formation are H+ consuming and hence

OH- producing reactions, the corresponding cathode would only shift to a more alkaline pH.

Hence, pH correction is needed to avoid cathode alkalinization during BES operation.

Figure 1.8 Change in the cathodic reduction potential versus change in cathodic pH. Values

calculated according to e.q. 1-9. ΔpH = pH at t>0 minus pH at t=0.

1.1.5.1.2 Conventional and State-of-the-Art Approaches for pH control in BES

The most commonly used pH control methods in the literature are the addition of buffer

(e.g. phosphate buffer) and by dosing acid/or base to the electrolyte (Rozendal et al. 2008a).

Although they could effectively address the pH limitation in laboratory studies, they are not

preferred for large scale wastewater treatment processes.

Recently, a complete loop-concept has been proposed and validated as a mitigation

strategy to overcome the aforementioned pH limitation during MFC operation (Freguia et al.

2007c). This is an interesting and could be a practical concept where the anodic effluent serves

as the influent for the cathode compartment in a two-compartment MFC. As the biological

y = -59x

-300

-200

-100

0

100

200

300

-5 -3 -1 1 3 5∆pH

∆Ecathode

(mV)

Alkalinization decreasing

Cathodic E

Acidification increasing

Cathodic E


- 24 -

oxidation of the electron donor substrate (here acetate) in their MFC anode is a H+ producing

reaction (see reaction 1-1 in Table 1.2), the anodic effluent becomes acidified (pH drops from

about 7.1 to 5.8 - 6.5). The hydraulic connection between the acidified anolyte and the catholyte

could effectively alleviate the alkalinization at the cathode, avoiding the need for pH control or

catholyte replacement (Freguia et al. 2007c). The same concept has been validated in a separate

report with a different BES configuration (Clauwaert et al. 2009).

However, similar to most other configurations reported in the literature the

aforementioned pH control strategy depend heavily on mechanical pumping or recirculation of

wastewater through the system (Clauwaert et al. 2009; Freguia et al. 2007c). Wastewater needs

to be continuously lifted up from the bottom inlet to the top of the anodic compartment, from

which the anodic effluent overflows and trickles over an air-exposed cathode. As the energy

required for lifting up a unit volume (i.e. a unit weight) of wastewater within the anodic

compartment is directly proportional to the compartment‘s height (see Figure 1.2), more energy

needs to be invested to lift the wastewater in a taller system. Such concern has not been

considered, and yet it may set a limit to the up-scaling potential for this kind of configuration.

1.1.5.2 Poor Cathodic Oxygen Reduction in MFC

Apart from the pH splitting limitation, the high overpotential associated with the cathodic

oxygen reduction using graphite and carbon electrodes often limits the MFC performance.

Although modifying the cathode with a platinum catalyst is a proven way to alleviate such

overpotential limitation, the use of platinum cannot be justified in MFC processes because of

costs and environmental impact in its production (Freguia et al. 2007b; Harnisch and Schroder

2009). Recent development of biocathode (bacterial catalyzed cathodic reduction) is seen as a

more sustainable and practical way to overcome the cathodic oxygen reduction problem

(Clauwaert et al. 2007b; He and Angenent 2006; Rabaey et al. 2008). However, the application

of biocathode still remains largely unexplored.


- 25 -

1.2 Aim and Scope of the Thesis

The overall aim of this PhD thesis is to explore the potential of BES for energy recovery

from organics in wastewater. The research program is broadly divided into two main themes:

1) Fundamental understanding of BES for electricity generation with emphasis on the

microbe-anode interactions. (Chapter 2 to 5)

2) Developing a practical reactor configuration to overcome the established bottlenecks.

(Chapter 6 to 8)

In general, acetate was used as a model substrate in this thesis as it is the most commonly

selected dissolved organic compound in many other wastewater treatment studies. Further,

acetate is considered as a dead-end product in dark fermentation because it cannot be further

converted into hydrogen due to thermodynamic and biochemical reasons (Cheng et al. 2007).

As mentioned earlier in this chapter (Chapter 1), one of the fundamental and common

steps in both MFC and MEC is the anodic oxidation of organic electron donors. Therefore,

Chapter 2 aims at establishing an anodophilic biofilm from an activated sludge (a commonly

used MFC mixed culture inoculums) in a two-chamber MFC. Since a non-limiting cathode was

essential to evaluate the anodic process (e.g. establishment of biofilm), ferricyanide was used as

the catholyte to offer a stronger and a more stable cathodic reaction.

The anode is an analogue to the terminal electron acceptor for the electrochemically

active bacteria to oxidize organic substrates in a BES. Hence, the energetics of the microbial

oxidation process largely depends on the potential of the anode. However, in the literature the

dependency of the microbial reaction on the anodic potential has not been investigated in detail.

Hence, Chapter 3 investigates the ―attractiveness‖ of the anode to the microbial rate of charging

it with electrons. While the effect of substrate concentration on the rate of microbially catalyzed

reactions is well understood and described as affinity, such relationship has not been described

for insoluble electron acceptors such as an anode. Such relationship was explored by controlling

the external resistance of the MFC.

Although online process control is used for many bioprocesses (e.g. sequencing batch

reactor for wastewater treatment) it is not established for MFC. One of the main features in this

thesis is the application of computer-feedback for BES process control. As an example to this,

Chapter 4 introduces a novel approach for controlling the biofilm-electrode potential by


- 26 -

feedback controlling the external resistance of an operating MFC. In analogy to a conventional

cyclic voltammetry study, this new approach could evaluate the redox behavior of an MFC

biofilm but without using a potentiostat. In this way, the MFC can still operate as a ―fuel cell‖

without being ―interrupted‖ by an external device (i.e. potentiostat) which normally does not

belong to the system. Such method could be useful for diagnosing system performance during

MFC power generation.

Cyclic voltammetry is a commonly used electrochemical method to study the electron

transfer between bacteria and the insoluble electrode in a BES. However, the complexity of a

biofilm-electrode system often limits the information obtained from this method. In Chapter 5,

a series of potentiostatic experiments was conducted to further explore the electron transfer

properties of an active anodophilic biofilm. Rather than using standard voltammetry techniques,

this thesis shows the merit of ultra slow scan rates (e.g. 0.01 mV·s-1

) for the study of

bioelectrocatalysis of an electrochemically active biofilm.

Chapter 6, 7 and 8 are devoted towards a practical and sustainable operation of BES for

energy recovery from wastewaters. The concept of using a single biofilm to alternately catalyze

the anodic oxidation of organics and a cathodic oxygen reduction is investigated in Chapter 6.

Its objectives are twofold: (1) to overcome the generic pH drifting problem inherent to the use of

membranes in BES and (2) to alleviate the oxygen reduction overpotential at a (non-catalyzed)

plain granular graphite electrode using an anodophilic biofilm.

Based on the findings of Chapter 6, a novel membrane-less BES configuration, known as

Rotatable Bio-Electrochemical Contractor (RBEC) was designed, fabricated and evaluated for

its performance as a MFC for electricity generation in Chapter 7; and as a MEC for hydrogen or

methane production in Chapter 8.

Finally, Chapter 9 discusses and summarizes the insights obtained in this thesis. Future

research directions and outlook are addressed.

- 27 -

2 Establishing an Anodophilic Biofilm in a MFC with a

Ferricyanide-Cathode and Online pH Control

Chapter Summary

Maintaining a suitable anodic condition in a bioelectrochemical system such as microbial

fuel cell (MFC) is essential for establishing active anodophilic biofilm from the initial inoculum

at an anode. In particular, a pH neutral operating anolyte and a non-limiting MFC cathode are

essential to such acclimation process.

This initial experimental chapter aims at establishing an anodophilic biofilm from a

mixed culture activated sludge inoculum in a pH-controlled, ferricyanide-cathode driven two-

chamber MFC. The MFC was inoculated with activated sludge (10%, v/v) and operated at fed-

batch mode at 30oC for 30 days. A synthetic wastewater amended with acetate (10 mM) and a

potassium ferricyanide solution (50 mM, 100 mM pH 7 phosphate buffer) were used as the

anolyte and catholyte, respectively. The external resistance of the MFC was predominately

maintained at low level (1-5 Ω) to avoid anodic limitation. Anolyte pH was controlled at 7 by

dosing NaOH (1M).

Regular polarization curve analysis indicated an overtime increase in MFC power output.

The rapid buildup of high power densities of the MFC (415 W·m-3

after five days) suggest that

an activated sludge mixed culture inoculum can readily become highly anodophilic at a MFC

anode, provided that factors such as electrolyte pH, external resistance and cathodic oxidizing

power are not limiting. However, the use of ferricyanide catholyte and the continuous alkaline

dosing for pH control are both non-sustainable and therefore unsuitable options for practical

BES application. More suitable alternatives should be explored for a sustainable MFC operation.

Chapter 2: MFC Start-up and Basic Understandings

- 28 -

2.1 Introduction

Start up of bioelectrochemical systems (BES) for simultaneous electricity production and

wastewater treatment usually involves the use of mixed cultures bacterial inoculums such as

activated sludge or anaerobic digester sludge (Logan 2005b; Rozendal et al. 2008a). Under

suitable acclimation condition, it is expected that the BES is able to select and enrich its own

electrochemically active consortia from the mixed culture over time (Kim et al. 2004a; Liu et al.

2008; Rabaey et al. 2003). During this process, the bacteria are able to establish their own

―redox link‖ with the insoluble anode without depending on artificially added redox mediators

(Jang et al. 2004; Schroder 2007). This ―link‖ allows the electrochemically active (or

anodophilic) bacteria to oxidize their electron donors (organics in a wastewater) and transfer the

electrons to the anode for their own energetic need. Such anodic electron transfer process

depends largely on the bacterial activity and the operating condition inside the anode chamber.

Thus, a suitable anodic condition is essential for the establishment of anodophilic biofilm in a

BES during the acclimation period.

One of the challenges in operating a MFC anode is to overcome the acidification of a

poorly pH-buffered anolyte over time (Gil et al. 2003; Pham et al. 2009). Such anolyte

acidification is caused by the proton-producing anodic oxidation of organics (e.g. acetate) (eq. 2-

1).

CH3COO- + 4 H2O 2 HCO3

- + 9 H

+ + 8 e

- (eq. 2-1)

As the electrons are accepted by the anode, the protons are concomitantly released to the anolyte.

Ideally, these released protons will be migrated from the anolyte to the catholyte via a cation-

exchange membrane to maintain the ionic charge balance in the BES. However, under the

prevailing operating environment in most wastewater treating BESs (low ionic strength, neutral

pH, low pH buffer capacity and ambient operating temperature), the proton flux is highly

inefficient. Other cations (e.g. sodium, magnesium or calcium ion, etc) will preferentially

migrate from the anolyte to the cathode (Rozendal et al. 2006a). As a consequent, the anolyte

becomes acidified. It is well accepted in the literature that acidification of the anolyte may

hamper both the electrochemical and biological activity of the anodic process (Clauwaert et al.

2008a; Harnisch and Schroder 2009; Picioreanu et al. 2009). Hence, maintaining a stable and

neutral anolyte pH is expected to allow stable anodic performance.


- 29 -

From a bacterial standpoint, the functions of an anode in a BES are twofold: (1) it can

serve as a physical surface for biofilm-forming bacteria to attach and grow; (2) it serves as the

terminal electron acceptor for the electrochemically-active bacteria to derive metabolic energy.

Although it is quite common that electrochemically active bacteria in BES tend to exist as a

biofilm attaching at the anode surface, this is not an essential prerequisite for anodic electron

transfers because suspended bacteria could also generate anodic current using soluble redox

mediators (Choi et al. 2003; Rabaey et al. 2007; Wang et al. 2007).

Addition of artificial exocellular redox shuttling compounds (mediators) was originally

thought to be crucial for an effective anodic electron transfer (Bennetto and Stirling 1983;

Tanisho et al. 1989). However, many recent studies have already shown that bacteria are able to

self-mediate such electron transfer either by directly using their outer-membrane redox

components (e.g. c-type cytochromes) or indirectly using self-excreted soluble redox

compounds (i.e. mediators) (Lovley 2006a; Rabaey et al. 2007; Reguera et al. 2006; Schroder

2007).

This chapter aims to establish an anodophilic biofilm from a mixed culture activated

sludge using a computer-pH controlled two-chamber microbial fuel cell (MFC). No artificial

mediator was provided during the acclimation process. The performances of a ferricyanide-

cathode and a dissolved oxygen based cathode are also compared.


- 30 -

2.2 Experimental Section

2.2.1 Microbial Fuel Cell Construction

A dual-compartment microbial fuel cell made of transparent Perpex was used in the

present study (Figure 2.1, Photo 2.1). It was kindly provided by Dr. Korneel Rabaey from the

Advanced Water Management Center, The University of Queensland.

Figure 2.1 A schematic diagram of the microbial fuel cell used in the present study.

The two compartments were physically separated by a cation selective membrane

(Ultrex™ CMI7000, Membranes International Inc.) with surface area of 168 cm2. The two

compartments were of equal volume and dimension (316 mL (14 cm x 12 cm x 1.88 cm)) (Photo

2.2A). Granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-6 mm diameter,

porosity of 45%, see Photo 2.2B) was used as both the anode and cathode electrode, which

caused a decrease of the internal liquid volume from 316 (14 cm x 12 cm x 1.88 cm) to 120 mL

(net anodic volume). The graphite electrodes were pretreated by soaking in 1 N HCl for 24 hours,

and were washed thoroughly with running deionised water before packing into the chambers.

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Stirrer

1M NaOH

Acetate Feed

Moisten

Air

Variable External Resistor

(PC-Controllable Relay Board)

Anolyte pH Anolyte Eh

Catholyte Eh

Ag/AgCl

Reference

Electrode

Biofilm-Anode

(Granular Graphite)

Counter Electrode

(Granular Graphite)

Cation Exchange Membrane

e- e-

Recirculation

Pump

MF

C V

olta

ge

Bio

film

-Ele

ctr

ode

Po

ten

tial

Catholyte

Graphite Rod

( Diameter =

5mm )

Out In

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Stirrer

1M NaOH

Acetate Feed

Moisten

Air




Catholyte Eh

Ag/AgCl

Reference

Electrode

Biofilm-Anode

(Granular Graphite)

Counter Electrode

(Granular Graphite)


e- e-

Recirculation

Pump

MF

C V

olta

ge

Bio

film

-Ele

ctr

ode

Po

ten

tial

Catholyte

Graphite Rod

( Diameter =

5mm )

Out In


- 31 -

Contacts between the external circuit and the granular graphite electrodes were made using

graphite rods (5mm diameter).

Photo 2.1 Photo of the microbial fuel cell set-up used in this study.

Photo 2.2 Photos of (A) the two-compartment reactor (ion-exchange membrane is not

shown) and (B) the graphite granules electrode material used in this study.

2.2.2 Bacterial Inoculum and Medium

To start up, the anode chamber was inoculated with activated sludge collected from a

wastewater treatment plant (Woodman Point, Perth, WA). 10% (v/v) of freshly collected

activated sludge with biomass concentration of 2.0 g L-1

(predetermined by using a dry-weight

method) was mixed together with synthetic wastewater. Composition of the synthetic

1 cm

B A


- 32 -

wastewater was (mg L-1

): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300,

FeSO4·7H2O 6.25, , and 1.25 mL L-1

of trace element solution, which contained (g L-1

):

ethylene-diamine tetraacetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24,

MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O

0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. Sterile concentrated yeast extract solution was

periodically added to the anolyte to support healthy microbial growth (50 mg L-1

). 100 mM of

phosphate buffer was also added to the anolyte in order to resist pH fluctuation and to improve

ionic strength of the anolyte.

2.2.3 Start-Up and Operation of a Ferricyanide-Cathode MFC

The MFC was operated in a fed-batch mode with both catholyte and anolyte

continuously re-circulating over the cathode and anode, respectively at a fixed flow rate of 8 L h-

1 (unless otherwise specified). Sodium acetate was used as the sole energy source for bacteria.

For replenishment of electron donors in the MFC, a designated amount of a concentrated sodium

acetate stock solution (1M) was injected into the anode chamber via a septum-secured injection

port just before the inlet of the anode chamber. Over the start up period, the anolyte (i.e. the

bacterial medium) was completely renewed in about every 3 days in order to remove suspended,

plank-tonic bacterial cells and any accumulated toxic end products from the system. Since the

present study is focused on the anodic process of MFC, a strong and stable cathodic reaction is

essential to maintain a strong oxidizing anode. The external resistance of the MFC was

predominately maintained at 1-5 ohm. Hence, potassium ferricyanide (50 - 100 mM),

K3Fe(CN)6 (Sigma-Aldrich, Inc., Purity ca. 99 %) was selected as the terminal electron acceptor

in the catholyte despite its non-sustainable nature compared to oxygen (air)-cathode (Freguia et

al. 2007a; Logan et al. 2006). The cathode chamber was continuously fed with 50 mM air-

saturated potassium ferricyanide, K3Fe(CN)6 solution with 100mM phosphate buffer to resist pH

changes (pH was adjusted to 7 with 5N NaOH). The ferricyanide solution was recycled over the

cathode and was running through a 2 L aeration glass bottle which contained a gas diffuser for

continuous purging of the solution with 0.45 µm-filtered airs at a flow rate of 1 L air min-1

. The

catholyte was renewed periodically to maintain a high and stable oxidation-reduction potential

for effective reduction reaction of the MFC. Loss of oxidizing power of the ferricyanide solution

is indicated by a decolorization of the solution (Photo 2.3). Designated flow rates in both

chambers were adjusted by using peristaltic pumps. Unless otherwise stated, the MFC was

operated at about 30oC and atmospheric pressure.


- 33 -

Photo 2.3 Loss of oxidizing power of the ferricyanide solution indicated by a de-colorization of

the solution (left: spent solution; right: fresh solution).

Control and monitoring of the microbial fuel cell was partially automated. Potential

differences between anode and cathode (i.e. voltage), Eh in both anolyte and catholyte, and pH

of anolyte were monitored continuously by using LabVIEW™ 7.1 software interfaced with a

National Instrument™ data acquisition card and all the data were logged into an excel

spreadsheet with a personal computer (data logging frequency was depending on the signal

changes at different experiments). Since acetate oxidation always resulted in the production of

protons which caused a decrease in pH to the levels that inhibited microbial activity as reflected

by decreased current production in the MFC (data not shown), pH of the anolyte was controlled

at 7 by adding 1-2 M NaOH using a computer-controlled peristaltic pump throughout the whole

study.

2.2.4 Start-up of Anodophilic Activity in a Potentiostatic-Coupled MFC

A separate experiment was conducted to verify results obtained from the startup of

anodophilic activity with the ferricyanide-cathode, the MFC was coupled with a potentiostat

such that any possible cathodic or ohmic limitations in the system would be eliminated. The

current generation would only depend on the anodic microbial oxidation process. A three-

electrode potentiostat was used to control the anode of the same MFC at a fixed potential. The

working, counter and reference electrodes of the potentiostat were connected to the anode,

cathode and the silver-silver chloride reference electrode of the MFC, respectively. The anodic

potential was controlled at -300 mV vs. Ag/AgCl throughout the experiment. This potential was

chosen because it is slightly higher than the optimal anode potential observed in the MFC


- 34 -

operated with a ferricyanide cathode in our lab (data not shown). Over the period, the anolyte

was regularly renewed (once per about 3 days). Coulombic recovery of acetate oxidation by the

anodic biofilm was tested (see below for details) by adding known amount of sodium acetate

into an acetate-starved anodic half cell.

2.2.5 Performance of a Dissolved Oxygen-based Catholyte

With the established anodic biofilm, the ferricyanide catholyte was replaced with an

aerated phosphate buffer solution in order to evaluate the performance of dissolved oxygen

(DO)-based cathode. The cathodic chamber of the MFC was repacked with fresh graphite

granules to assure the absence of any residual ferricyanide in the cathode. The catholyte (total

volume of 1L) was continuously purged with humidified air (air flow rate was 1L·h-1

) in an

external aeration container to compensate the lost of dissolved oxygen after running through the

cathode chamber. The DO concentrations (DO, mmol O2 L-1

) of the catholyte at the aeration

bucket (DOin) and outlet (DOout) of the cathodic chamber were continuously monitored with two

identical DO probes (Mettler Toledo, InPro6800; 4100 PA D/O Transmitter), respectively.

The hydraulic retention time (HRTcath.) of the catholyte (27 sec, i.e. 0.0075 h) in the

cathodic chamber was obtained from the catholyte recirculation rate (here 16 L∙h-1

). The

apparent cathodic oxygen reduction rate, ORR (mmol O2∙L-1

∙h-1

) could thus be calculated

according to eq. 2-2.

.

1- 1-

2

)()hLO (mmol ORR

Cath

inout

HRT

DODO (eq. 2-2)

2.2.5.1 Effect of Catholyte pH and Phosphate Buffer Concentration

The effect of catholyte pH on the open circuit voltage (i.e. electromotive force) was

investigated. The anode of the MFC was saturated with 10 mM acetate and the anolyte pH was

maintained at 7 by NaOH dosing. A series of phosphate buffer (50mM) solutions at pH ranging

from 0.5 to 11 were loaded into the cathodic chamber. Steady state performance of the MFC was

recorded at each pH level.


- 35 -

The effect of phosphate buffer concentration of the catholyte on MFC performance was

evaluated with the DO-based cathode. A series of catholyte solutions with different phosphate

buffer concentrations (from 0 to 200 mM) was prepared (pH 7). These buffer solutions had ionic

strength ranged from 0 to 1200 mM. The ionic strength (Ic) at each buffer concentration was

calculated according to the following equation (eq. 2-3):

2

12

1i

n

i

ic ZCI

(eq. 2-3)

Where Ci is the molar concentration (mM) of ion i, Zi is the valence of that ion species, and the

sum is taken over all ions in the solution. The relationship between ionic strength of the

catholyte and the phosphate buffer concentrations is shown in Figure 2.2.

Figure 2.2 Dependency of catholyte ionic strength on phosphate buffer concentration in the

catholyte (pH 7).

y = 6.00x - 0.00

R2 = 1.00

0

200

400

600

800

1000

1200

0 50 100 150 200

Phosphate Buffer Concentration (mM)

Ionic

Str

ength

(m

M)


- 36 -

2.2.6 Calculation and Analysis

2.2.6.1 Determination of Voltage, Current and Power Generation

MFC power output (P) was calculated from the voltage measured across a known

resistor (adjusted by using a variable resistor box). The current (I) was calculated according to

Ohm‘s Law, I = V/R, where V is the measured voltage and R (Ω) is the external resistance.

Power output was calculated according to: P=V x I. Since it was impossible to determine the

projected surface area of the granular graphite electrode used in this study, current and power

density could only be calculated by normalizing the values by the working volume of the anodic

chamber (i.e. 150 mL). Internal resistance (Rint) of the MFC was determined according to the

polarization slope method reported elsewhere (Logan 2008) (Refer to Appendix 2 for details)..

Coulombic efficiency was calculated by dividing the number of electrons transferred to

the anode to generate current by the number of electrons contained in the amount of acetate

being oxidized in the MFC (determined by GC-FID).Current (in milliamperes, mA) was

converted to electrons recovered by using the following conversions: 1 Coulomb (C) = 1 Amp x

1 second, 1 C = 6.24 x 1018

electrons, and 1 mol = 6.02 x 1023

electrons (96,485 C/mol).

2.2.6.2 Polarization Curve Analysis

The MFC performance was examined by conducting current-voltage analysis or the so-

called polarization curve analysis in the first month (at day 5, 14, 22 and 30). This technique is

commonly used in many other studies for generating polarization curve analysis (Logan et al.

2006). This was done by a manual step decreasing of the external resistance. Before the analysis,

the anolyte was saturated with acetate (20 mM) and a fresh ferricyanide solution was used as the

catholyte. The MFC was allowed to equilibrium at open circuit for about 2 hour to obtain the

maximal cell voltage. Thereafter, the external resistance was decreased in a step-wise manner to

1 ohm. Only steady state parameter values as defined by Logan et al. (2006) were taken to

generate the final plots. Volumetric current and power densities were obtained by normalizing

the current and power with the net anodic volume.

2.2.6.3 Acetate Analysis

A Varian Star 3400 gas chromatograph (GC) fitted with a Varian 8100 auto-sampler was

used to quantify the acetate concentration of liquid samples collected from the anodic chamber.


- 37 -

The samples were subjected to centrifugation at 13,500 rpm for 5 min to separate any suspended

solids from the liquid phase. The supernatant was acidified with 1% (v/v) formic acid before

being injected (1 µL) into an Alltech ECONOCAPTM

ECTM

1000 (15 m × 0.53 mm, 1.2 µm i.d.)

column. The carrier gas (N2) was set at a flow rate of 5 mL min-1

. The oven temperature was

programmed as follows: initial temperature 80 oC; temperature ramp 40

oC min

-1 to 140

oC, hold

for 1 min; temperature ramp 50 oC min

-1 to 230

oC hold for 2 min. Injector and detector

temperatures were set at 200 and 250 oC respectively. The peak area of the Flame Ionisation

Detector (FID) output signal was computed via integration using STAR Chromatography

Software (© 1987-1995).


- 38 -

2.3 Results and Discussion

2.3.1 Quick Start-up of MFC using Ferricyanide-Cathode and Continuous pH-Static

Control

The described two-chamber MFC was inoculated with activated sludge, and operated for

30 days using ferricyanide as the catholyte. Over this acclimation period, the external resistance

of the MFC was predominately maintained at 1-5 ohms.

Polarization curve analysis was conducted regularly over time (Figure 2.3). It allows the

determination of both the maximal current and power densities as well as the electrochemical

properties of the anode and cathode. In general, the volumetric current and power densities

(normalized to net anodic chamber volume) of the MFC increased over time (Figure 2.3), and

the internal resistance of the MFC had decreased over time, reaching the lowest internal

resistance of only 0.54Ω at day 30 (Table 2.1). With a similar MFC set-up, Dr. Korneel Rabaey

(Advanced Water Management Center, The University of Queensland) had also recorded a

similar internal resistance of 0.61Ω by using electrochemical impedance spectroscopy method

(personal communication on 14th

, Sep 2007).

The overtime decrease in internal resistance is indicative for a well performing

anodophilic microbial community and demonstrates the effective design and operation of the

reactor. The bacterial culture in the anode chamber has become more capable in transferring

their electrons to the anode during the acclimation.

Table 2.1 A summary of the MFC performance over a 30-day development of the anodophilic

biofilm in the ferricyanide cathode-MFC.

Time Maximal Power

Density Maximal Acetate Oxidation Rate

External Resistance at Maximal Power

Density Internal Resistance*

Day W·m-3

mM·h-1

Ω Ω

5 415 1.2 6 2.26

14 868 3.0 2 1.20

21 1321 5.2 1 0.97

30 1735 6.1 1 0.54

Note: the maximal acetate oxidation rates are calculated assuming a 100% Coulombic conversion

of acetate into current (8 e- mole per acetate); *internal resistances were calculated by using

polarization slope method (Logan 2008, page 53-54) (see Appendix 2).


- 39 -

Figure 2.3 The power density (A) and polarization curves (B, cell voltage; C, electrode potential)

of the MFC at day 5 ( / ), day 14 ( / ), day 22 ( / ) and day 30 ( / ) after

inoculated with activated sludge (10%, v/v).

0

200

400

600

800

Cell

Voltage

(mV

)

0

500

1000

1500

2000

Pow

er

Density

(W m

-3,

Anodic

Void

Vol.)

.

-600

-400

-200

0

200

400

0 1000 2000 3000 4000

Current Density (A m-3

, Anodic Void Vol.)

Ele

ctr

ode P

ote

ntial

(mV

vs.

Ag/A

gC

l)

A

B

C


- 40 -

The power density obtained at day 5 (415 W·m-3

) has already approached the maximal

levels observed in other MFCs that have been operated for a longer acclimation period

(Aelterman et al. 2008; Kim et al. 2007). For instance, with a similar reactor geometry, ion-

exchange membrane and electrode materials, Aelterman et al. (2008) observed a maximal MFC

power density of 517 W·m-3

(also normalized to the net anodic volume) but after a 25 days

acclimation period. Instead of activated sludge, an effluent taken from an active acetate-fed

MFC was used as their start up inoculum. Overall, the result obtained here suggests that quick

establishment of anodophilic activity from an activated sludge could be achieved by using a

ferricyanide-cathode and continuous neutral pH static control.

2.3.1.1 The Cathodic Oxidation Power could influence the Anodic Microbial Activity only

at a Low External Resistance

The use of a ferricyanide-cathode had offered a strong oxidizing surface (anode) to the

electrochemically active bacteria for the efficient oxidation of acetate during the acclimation.

However, the rate of acetate degradation related strongly to the oxidation power of the catholyte

(i.e. catholyte Eh) only when the external resistance was low (here at 1 Ω) (Figure 2.4).

Figure 2.4 Effect of ferricyanide-catholyte redox potential (Eh) on the anodic acetate oxidation

rate at different external resistances. Acetate oxidation are calculated assuming a 100%

Coulombic conversion of acetate into current (8 e- mole per acetate). In all cases, the

anode was always saturated with acetate (10 mM) and catholyte pH remained at 7.

0

1

2

3

4

5

0 50 100 150 200

Catholyte Eh (mV vs. Ag/AgCl)

Ace

tate

Oxid

atio

n R

ate

(mM

·h-1

)

1Ω

5Ω

10Ω

100Ω

1MΩ

Weak Oxidant

Strong Oxidant


- 41 -

Such correlation seems to obey with the well known Michaelis-Menten type kinetics

behavior: where increasing oxidation power of the catholyte (catholyte Eh) increases the

substrate oxidation rate up to a point where further increase in catholyte Eh could not further

increase the substrate oxidation rate (Figure 2.4, 1 and 5 Ω). While at external resistances of 10

Ω or higher, increase in the oxidation power of the ferricyanide catholyte had no significant

effect on the anodic acetate oxidation rate.

Similar observation has been reported by You and coworkers (2006). With low

resistance between anode and cathode, increase in concentrations of the terminal electron

acceptors (permanganate was used in their study) in the cathodic chamber of the MFC could

increase current density whereas no significant effect was observed when higher resistance was

used.

Overall, the use of a very low external resistance (less than 5 Ω in the described MFC)

was beneficial or perhaps crucial to enable effective anodic oxidation by the electrochemically

active bacteria during the MFC operation.

2.3.2 Potentiostatically-Controlled MFC also revealed Quick Anodophilic Activity Onset

from the same Activated Sludge Inoculum

If the observed fast establishment of anodophilic activity was attributed to the provision

of a strong oxidizing surface (anode) to the bacteria (via ferricyanide with low external resistor),

other methods that can maintain a similar oxidizing capacity of the anode should also allow a

quick start up of anodophilic activity from the activated sludge under similar anodic condition.

One of these methods is by coupling the MFC with a potentiostat. Using potentiostats for

the evaluation of anodic processes in BES has been proposed as an appropriate method to rule

out system limitations such as ohmic resistance and cathodic limitations (Fricke et al. 2008;

Marken et al. 2002). Hence, a potentiostatic experiment was conducted to establish anodophilic

activity from the same activated sludge inoculum under a similar anolyte condition (pH 7, 30oC).

The anodic potential was controlled at a constant level (here -300 mV vs. Ag/AgCl). The

ferricyanide at the cathode was replaced with a 50 mM phosphate buffer (pH7) (Figure 2.5).


- 42 -

Figure 2.5 Evolution of (A) current and (B) anolyte Eh after inoculation of an activated sludge

(10% v/v) in a pH-controlled potentiostatic bioelectrochemical cell (the same as used in

Figure 2.2). Initial acetate concentration in the medium was about 15 mM. pH was

controlled at about 7 by feedback controlled addition of a 1M NaOH solution (C). Values

reported in A and B are the means taken in 30 min intervals. Dotted arrows indicate

additions of 50mM sodium acetate.

2.3.2.1 Evolution of Anodic Current after a short Lag-phase of only One Day

After the bacterial inoculation, the anolyte Eh decreased almost linearly from about -50

to a steady level of about -430 mV in 20 hours (Figure 2.5B). Such a decline in Eh indicates that

the activated sludge could reduce their surrounding environment (anolyte) to a highly reduced

-500

50100150200250300350400450500550

Curr

ent (m

A)

-450

-350

-250

-150

-50

An

oly

te E

h (

mV

vs. A

g/A

gC

l) .

4

7

10

0 1 2 3 4 5 6 7 8 9Time (d)

pH

A

B

C

Medium renewal

Medium renewal

CE = 95.7%

(1)

(2)

(3)

(4) (5)

(6)


- 43 -

condition by metabolizing the acetate (electron donor). Such anaerobic condition would be

essential for the effective anodic electron transfer by the bacteria.

After a lag phase of only one day anodic current began to evolve exponentially,

signifying the onset of exponential growth of anodophilic active microorganism at the anode

(Figure 2.5A) (Hu 2008; Parot et al. 2007). Upon acetate depletion (data not shown), the anodic

current returned to the background, low level. This result demonstrates the ease of establishing

an anodophilic biofilm from an initially electrochemically inactive activated sludge inoculum.

2.3.2.2 Anodophilic Bacteria: Biofilm instead of Suspended Cells

At about day 3, the entire anolyte was discarded and the anodic chamber was flushed

with one liter (about 6 times the net anodic chamber volume) of fresh medium to remove any

plank-tonic cells or suspended cell debris before resuming the operation with an acetate-

saturated (50 mM) fresh medium (Figure 2.5A). Such flushing treatment did not retard the

anodophilic activity of the bacteria, but indeed a sharp increase in anodic current was observed

almost immediately after the medium renewal. Another flushing and medium renewal at about

day 7 resulted in a similar effect on anodic current (Figure 2.5A). Hence, the observed anodic

current was not due to the suspended bacteria but to the metabolizing biofilm attached to the

anode. This observation has also been reported by others using activated sludge as the source of

anodophilic bacteria (Lee et al. 2006; Yoon et al. 2007).

2.3.2.3 The Anodophilic Biofilm Established in the Potentiostatically Controlled MFC

also gave similar Power Output as the Ferricyanide-Cathode MFC

To compare the anodophilic biofilms established from the ferricyanide-cathode MFC and

the potentiostatically controlled MFC on a power production basis, the current obtained from the

latter was further processed to obtain an extrapolated hypothetical power output (Figure 2.6).

Assuming that the biofilm-anode was coupled to a ferricyanide-cathode, the theoretical MFC

cell voltage can be deduced from the difference between the poised anodic potential (here -300

mV vs. Ag/AgCl) and the cathodic potential at the corresponding current level. With this

estimated voltage, the extrapolated power output was calculated according to Power = Voltage ×

Current.


- 44 -

Figure 2.6 Extrapolated MFC power output using the anodophilic biofilm established from a

potentiostatic cell. Three power outputs values are depicted here. They are obtained from

peak 1, 3 and 5 in Figure 2.5. Data of the ferricyanide cathode are taken from Figure

2.3C (day 30). Cath. E = cathodic potential; An. E = anodic potential.

Similar to the ferricyanide-cathode MFC, the extrapolated power output of the

potentiostatically established anodophilic biofilm also increased over time (Figure 2.6). The

power outputs of the potentiostatically controlled MFC are in general higher than that of the

ferricyanide-cathode MFC. For instance, at day 5 the power densities of the ferricyanide-cathode

MFC and the potentiostat-MFC were 415 and 864 W·m-3

, respectively. This may be due to the

slightly more favorable operating anodic potential (-300 mV compared to <-400 mV))

maintained by the potentiostat. Nevertheless, it can be concluded that the anodophilic biofilms

established by using these two different approaches were similar.

Overall, this estimation suggests that even a young (ca. one week) anodophilic biofilm

could easily enable a MFC power output beyond a benchmark target of 400 W m-3

(as proposed

by Clauwaert et al. (2008a)) if both the cathodic reaction and electrolyte pH were not limiting

during the acclimation phase.

0

100

200

300

400

500

-400 -200 0 200

Electrode Potential (mV vs. Ag/AgCl)

Curr

ent

(mA

)Bioanode Potential(Poised at -300 mV)

Cathodic Potential(Ferricyanide-cathode)

Day 2.3 = 457 W·m-3

Day 5 = 864 W·m-3

Day 7.4 = 904 W·m-3

Hypothetical Power Output:

Power = (Cath.E – An.E)×Current


- 45 -

Figure 2.7 Scanning electron micrographs of the granular graphite anode before (D, E and F)

and after 200 days (A, B and C) in an acetate-fed microbial fuel cell anodic chamber.

2.3.2.4 SEM reveals only Thin and Low Biomass Density at the Highly Active Biofilm-

Anode

After operating the MFC for about 200 days, scanning electron microscopy photos were

taken to examine the morphology of the biofilm attached on the granular graphite anode (Figure

2.7). Surprisingly, only a thin, open structured and low biomass density biofilm was observed at

the electrode surface. Examination of replicates samples revealed similar morphology. This

implies that the established anodophilic biofilm has a very high specific activity (current

generation per amount of biomass).

The observation here seems to corroborate with the results reported by Altermam et al.

(2008). In their active MFC anodic chambers, these authors had determined a maximal biomass

density of 0.58 g VSS∙L graphite -1

, which they claimed is approximately 30 times lower

compared to biomass concentrations in anaerobic digesters (10-20 g VSS∙L-1

) and yet the

specific biomass activity of their biofilm (3.4 g COD g∙VSS-1

∙d-1

) was higher than that of

anaerobic systems (about 2 g COD g∙VSS-1

∙d-1

) (Aelterman et al. 2008; van Lier et al. 2008)

(Table 2.2).

If we assumed the anodic chamber of the described MFC in this study also has the same

biomass density of 0.58 VSS∙L graphite -1

, the specific biomass activity obtained at the maximal

10 µm 5 µm

2 µm10 µm 5 µm

2 µm10 µm10 µm10 µm 5 µm5 µm5 µm

2 µm2 µm2 µm10 µm10 µm10 µm 5 µm5 µm5 µm

2 µm2 µm2 µm

A B C

E D F


- 46 -

power output at day 30 would be 17.4 g COD g∙VSS-1

∙d-1

, which is about 6 times higher than the

typical activity of other MFC reported in the literature and is about 35 and 9 times higher than

that of activated sludge and anaerobic digestion treatment, respectively (Table 2.2) (Pham et al.

2009). This distinctive feature may give MFC a competitive advantage over other wastewater

treatment processes especially activated sludge processes as the post-treatment cost of the

excessive sewage sludge is tremendous.

Table 2.2 Comparison of parameters characterizing MFC anode, aerobic activated sludge and

anaerobic digestion for wastewater treatment.

Wastewater Treatment Technology

Yield Coefficient (Y)

Biomass Concentration (per reactor volume)

COD Removal Rate

Specific Biomass

Activity

g∙VSS-1∙g COD

-1 g∙VSS

-1∙L

-1 kg COD m

-3 d

-1 g COD g∙VSS

-1∙d

-1

a Aerobic Activated

Sludge 0.4 2 - 4 0.5 - 1 0.1 0.5

a Anaerobic Digestion 0.05 – 0.1 10 - 20 10 - 20 1 - 2

a MFC Anode

Literature 0.02 – 0.23 0.25 - 0.58 0.3 – 1.5 1.51 - 6.93

MFC Anode This Chapter

0.1 a 0.4

a 3.0 - 9.0

b 6.5 - 17.4

c

a Values after Pham et al.(2009);

b Values are estimated from current (100% Coulombic efficiency was

assumed); c Values are estimated based on a current output of the described MFC between 150 and

450 mA.

2.3.3 Performance of a Dissolved Oxygen-based Cathode

In this study, the use of ferricyanide enabled the effective start up of anodophilic activity

and the high electricity output of the MFC. However, ferricyanide is not sustainable and hence it

has very limited merit for wastewater treatment application. Oxygen, on the other hand, is

considered as the most sustainable electron acceptor in MFC cathode because of its virtually

unlimited availability from the atmosphere, and in principle it has the highest redox potential

(Freguia et al. 2007b; Oh et al. 2004). However, it is well accepted that the imperfect catalysis of

the oxygen reduction reaction (eq. 2-4) at plain, non-catalyzed electrode surface limits the

cathodic reaction and hence the MFC performance.


- 47 -

O2 + 4 H+ + 4 e

- 2 H2O (eq. 2-4)

With the same established anodic biofilm, the ferricyanide catholyte was replaced with

an aerated phosphate buffer solution to evaluate the performance of a dissolved oxygen (DO)-

based cathode. In particular, the effects of catholyte pH and phosphate buffer were evaluated.

2.3.3.1 Acidified Catholyte increases Open Circuit Voltage of the MFC

Acidifying the catholyte (50 mM aerated phosphate buffer) could increase the open

circuit voltage (OCV) of the MFC (Figure 2.8A). Such effect was due to an increase in the

cathodic potential (Figure 2.8B). From a thermodynamic perspective, the electrode potential in a

MFC is linearly dependent on the local pH of the electrolyte (refer to Figure 1.8 in Chapter 1).

Decrease in catholyte pH by one unit should increase the cathodic potential by 59 mV. However,

such dependence did not hold true here, especially at a higher pH range (>pH7). For example,

when the catholyte pH dropped from about 11 to 9.3, the catholyte potential increased from -67

to -64 mV (2 mV/pH). When the catholyte pH changed from 3 to 0.5, the catholyte potential

increased from +170 to +360 mV (76 mV/pH) (Figure 2.9).

Figure 2.8 (A) Plots of cell voltage; cathodic potential and anodic potential as a function of

catholyte pH under open circuit operation. (B) Correlation between the open circuit

potential and the cathodic potential in A. All catholyte contained 50 mM phosphate

buffer. All values were recorded at equilibrium state. Anolyte was saturated with 5 mM

acetate, pH 7. The ferricyanide cathode data in A was obtained from Figure 2.3B.

Catholyte pH

Ce

ll V

olta

ge

(m

V) o

r E

lectr

od

e

Po

ten

tia

l (m

V v

s. A

g/A

gC

l)

-600

-400

-200

0

200

400

600

800

0 2 4 6 8 10 12

Cell Voltage

Cathodic Potential

Anodic Potential

OCV of Fe(CN)63-Cathode

y = 1.009x + 520.940

R2= 0.999

0

200

400

600

800

-100 0 100 200 300 400

Cathodic Potential (mV vs. Ag/AgCl)

Op

en

Circu

it P

ote

ntia

l (m

V)

(A) (B)


- 48 -

Figure 2.9 Change in cathodic potential per unit of pH as a function of the catholyte pH.

(plotted from the data in Figure 2.8A).

Similar observation was noted twice by Zhao and coworkers from their experiments

using graphite electrodes modified with various chemical catalysts. They attributed the

phenomena to a change in the nature of the rate-determining step in the overall cathodic reaction,

most likely from a 4 e-/ 4H

+ (or 2 e

-/ 2H

+) to a 4 e

-/ 2H+ (or 2 e

-/ 1H

+) reaction (relative to one

oxygen molecule) (Zhao et al. 2005, 2006a).

Whether or not the observed discrepancy was due to the same reason is beyond the scope

of the present study. Yet, the significance of the result here is that when compared with a

ferricyanide cathode, which always gave an OCV of about 750 mV the oxygen-based cathode

could give similar OCV only at catholyte pH of less than 2 (Figure 2.8A). Such an acidic

condition is unlikely to be realistic under normal MFC operation, especially in systems designed

for wastewater treatments. On the other hand, the cathodic reduction of ferricyanide does not

involve protons (eq. 2-5), and hence varying the catholyte pH does not affect the cathodic

potential (Zhao et al. 2006b). Based on this (but ignore the sustainability concern), ferricyanide

can allow a more stable cathodic reduction reaction for the MFC compared to a DO based

cathode.

Fe(CN)63-

+ e- → Fe(CN)6

4- (eq. 2-5)

2.3.3.2 Phosphate Buffer improves the Dissolved Oxygen-Cathode Performance

Increasing the phosphate buffer concentration in the catholyte could increase the

0

20

40

60

80

0 2 8 10 12Catholyte pH

Cath

odic

Pote

ntial C

hange

mV

/pH

Theoretical Change

59 mV/pH


- 49 -

performance of the DO-cathode (Figure 2.10). When DO was not limiting (>5 mg L-1

), a

Monod-type relationship between the MFC current and the buffer concentration was observed:

maximal current ≈ 40 mA; half saturation buffer concentration ≈ 8 mM, and saturation buffer

concentration ≈ 100 mM (Figure 2.10A). This result indicates that the cathodic oxygen reduction

depends on the phosphate buffer concentration in the catholyte (Figure 2.10C).

Figure 2.10 Effect of phosphate buffer concentration in catholyte on (A) voltage; current; power

density; and (B) electrode potentials of the acetate-fed microbial fuel cell; (C) cathodic

oxygen reduction rate as a function of phosphate buffer concentration in the catholyte.

External resistance 1.5 Ω. Anolyte was saturated with about 5 mM acetate. pH of both

anolyte and catholyte was 7, temperature 30oC. Catholytes were always saturated with

>5 mg·L-1

dissolved oxygen.

The positive effect of phosphate buffer was possibly due to its capacity to improve the

proton availability for the cathodic reaction, which consumes protons (eq. 2-4). At pH 7, proton

concentration is very low, i.e. 0.001 mM. With a poorly pH-buffered catholyte, proton

consumption at the cathode would likely be faster than the proton diffusion from the bulk to

compensate for the proton loss at the cathode. Consequently a localized pH increase is

0

10

20

30

40

50

60

70

Cell Voltage

Current

Power Density

-520

-500

-480

-460

-440

0 200 1000 1200

Cathode

Anode

(B)

(A)

0 50 100 200150

Ionic Strength (mM)

Vo

lta

ge

(m

V);

Cu

rre

nt (

mA

);

Po

we

r D

en

sity (W

·m-3

)

Ele

ctr

od

e P

ote

ntia

l(m

V v

s. A

g/A

gC

l)

Phosphate Buffer (mM)

0 200 1000 1200Ionic Strength (mM)0

1

2

3

4

5

Oxyg

en R

eduction R

ate

(mm

ol∙O

2∙L

-1∙h

-1)

0 50 100 200150

Phosphate Buffer (mM)

(C)


- 50 -

established at the cathode, shifting the cathodic potential to a more negative value and hence the

cell voltage and current are diminished (Zhao et al. 2006b).

The situation would be very different if the bulk catholyte contained a sufficient amount

of phosphate buffer. For instance, at a phosphate buffer concentration of 100 mM, the diffusion

rate of the phosphate buffer species in the catholyte could be about 105 times higher than that of

protons, even though the diffusion coefficient of free protons is about 10 times higher than that

of buffering species (monobasic and dibasic phosphate: 1.0 × 10-5

and 0.86 × 10-5

cm2·s

-1 at

30oC (Fan et al. 2007; Mason and Culvern 1949)). Hence, the buffer-mediated transfer of

protons from the bulk catholyte to the cathodic reaction site could be efficient enough to meet

the proton demand in the cathode oxygen reduction, resulted in an improved cathode and MFC

performance.

Nevertheless, there are at least two drawbacks that may disfavor the use of phosphate

buffer in BEC for wastewater treatment: 1) huge amount of buffer is required especially for

large scale wastewater treatment, production and energy cost of the buffer would be huge; 2)

discharge of the buffer-containing effluent would potentially cause subsequent pollution at the

receiving water bodies, e.g. eutrophication. Hence, a more sustainable alternative of

compensating the proton demand of the oxygen-based cathodic reaction has to be developed.

2.3.3.3 Ferricyanide-Cathode Outperforms Dissolved Oxygen-Cathode

In spite of the beneficial effect of phosphate buffer on the cathodic oxygen reaction, the

maximal current of the DO-cathode MFC was only about 40 mA, which was about 8 times

lower compared to the current obtained from the ferricyanide-cathode MFC (Figure 2.11).

The inferior performance of the DO-cathode MFC was largely due to the huge

overpotential of the cathode (Freguia et al. 2007b; Rismani-Yazdi et al. 2008). The cathodic

potential of the DO-cathode was about 12 times lower (more negative) compared to ferricyanide

cathode (Figure 2.11), suggesting that the cathodic oxygen reduction is associated with a large

overpotential compared to ferricyanide.

Similar to others, the abiotic cathodic oxygen reduction was the limiting step to the

current generation in the described MFC (Clauwaert et al. 2007b; Freguia et al. 2007b; Oh et al.

2004; Rismani-Yazdi et al. 2008). Further studies are required to improve the cathodic oxygen

reduction in a sustainable manner.


- 51 -

Figure 2.11 Comparison between ferricyanide- and dissolved oxygen- based cathode in an

acetate-fed microbial fuel cell. External resistances of both settings were 1.5 Ω. Anolyte

was saturated with 5 mM acetate, pH7, 30oC. Both catholytes were amended with 100

mM phosphate buffer (pH7). Data of the ferricyanide cathode are taken from Figure 2.3C

(day 30). CP/ AP = cathodic/ anodic potential (mV vs. Ag/AgCl).

2.4 Conclusion and Implication

This chapter demonstrates that quick establishment of anodophilic activity from an

activated sludge could be achieved by using a ferricyanide-cathode and continuous neutral pH

static control of the MFC anode. Both the volumetric current and power densities of the MFC

increased over time. The overtime decrease in internal resistance is indicative for a well

performing anodophilic microbial community and demonstrates the effective design and

operation of the reactor. The bacterial culture in the anode chamber has become more capable in

transferring their electrons to the anode during the acclimation period. The power density

obtained at day 5 (415 W·m-3

) had already approached the maximal levels observed in other

similar-scale MFCs that have been operated for a longer acclimation period.

The developed anodophilic biofilm appeared to have a very high specific activity in

converting the soluble COD (here acetate) into electricity. This distinctive feature may give

MFC a competitive advantage over other wastewater treatment processes especially activated

sludge processes as the post-treatment cost of the excessive sewage sludge is tremendous.

-500

0

500

1000

1500

Current

(mA)

CP(mV) AP (mV)

Cell

Voltage(mV)

Power

Density(W·m-3)

Ferricyanide

Dissolved Oxygen


- 52 -

As demonstrated in this study, ferricyanide (Fe (CN)63-

) was a very effective terminal

electron acceptor for the cathodic process. It offered a stable reduction reaction even at the non-

catalyzed granular graphite cathode surface. However, due to the slow reoxidation of the

reduced state (ferrocyanide, Fe (CN)64-

), the regular catholyte renewal was essential to sustain

stable MFC performance. This requirement leads to the generation of a tremendous amount of

spent non-functioning ferrocyanide solution (chemical waste), leading to a cost and

environmental impact of the process.

Oxygen, on the other hand, is considered as the most sustainable electron acceptor in

MFC cathode because of its virtually unlimited availability from the atmosphere. It was

demonstrated that the addition of phosphate buffer could enhance the cathodic oxygen reduction,

most likely due to a facilitated buffer-mediated proton transfer from the bulk catholyte to the

cathode. However, the performance of the dissolved oxygen based cathode was insignificant

compared to that of a ferricyanide-cathode. Hence, future study should aim at improving the

oxygen-based cathodic reaction. Lastly, the continuous pH control by NaOH addition and the

regular replenishment of catholyte used in this study may facilitate the ion transfer in the system

in a non sustainable manner, allowing Na+ instead of H

+ transfer as the dominant mechanism of

ionic charge transfer. From a practical standpoint, a more sustainable pH control method has to

be developed.

- 53 -

3 Affinity of Microbial Fuel Cell Biofilm for the

Anodic Potential

(This chapter has been published in Environ. Sci. Technol. (2008) 42: 3828-3834)

Chapter Summary

In Chapter 2, the relationship between the anodic acetate oxidation rate and the catholyte

oxidation power (catholyte Eh) seemed to exhibit a well known Michaelis-Menten type kinetics

behavior: where at low external resistance (<5Ω) increasing catholyte Eh increases the substrate

oxidation rate until reaching a point beyond which further increase in catholyte Eh could not

further increase the substrate oxidation rate.

In analogy to the well established dependency of microbial reactions on the redox

potential of the terminal electron acceptor, the dependency of the microbial activity in a highly

active microbial fuel cell on the potential of the electron accepting electrode (anode) in a

microbial fuel cell (MFC) is investigated in this chapter. An acetate-fed, pH controlled MFC

was operated for over 200 days to establish a highly active MFC anodic biofilm using

ferricyanide as the catholyte and granular graphite as electrode material. From the coulombic

efficiency of 83% of the MFC the microbial activity could be recorded by on-line monitoring of

the current.

The results suggest that:

(1) in analogy to the Michaelis-Menten kinetics a half saturation anodic potential (here

termed kAP value) could be established at which the microbial metabolic rate reached

half its maximum rate. This kAP value was about -455 mV (vs. Ag/AgCl) for our

acetate driven MFC and independent of the oxidation capacity of the cathodic half cell;

(2) a critical AP (here termed APcrit.) of about -420 mV (vs. Ag/AgCl) was established that

characterizes the bacterial saturation by the electron accepting system. This critical

potential appeared to characterize the maximum power output of the MFC.

This information would be useful for modeling and optimization of microbial fuel cells and

the relative comparison of different microbial consortia at the anode.

Chapter 3: Affinity of MFC Biofilm for the Anodic Potential

- 54 -

3.1 Introduction

Microbial fuel cell technology (MFC) is an emerging environmental technology for

energy recovery from organic waste streams. Over the past few years, our knowledge on MFC

has been gaining momentum. Yet our understanding of the electron transfer from the bacteria to

the anode is still not complete (Kim et al. 2007; Lovley 2006b; Rabaey et al. 2007; Schroder

2007). In MFC an electron flow is generated between the anodic half cell in which the bacteria

form a biofilm on the anode surface and the cathodic half cell which uses a chemical electron

sink such as oxygen or ferricyanide (Logan et al. 2006). In the anodic half cell bacteria oxidize

organic electron donors such as acetate via the tricarboxylic acid (TCA) cycle and transfer the

gained electrons to internal electron carriers such as NAD+/NADH. A suitable external electron

acceptor is needed for the bacteria to recycle (re-oxidise) their internal electron carriers. Such an

acceptor may either be present extracellularly in the environment or generated intracellularly by

the bacteria (De Graef et al. 1999; Stams et al. 2006). In MFC the bacteria manage to transfer

the internally generated electron flow to a suitable solid conductive surface (anode). As the

anode receives the electrons its potential decreases to a lower level than that of the cathode in

the adjacent cathodic half cell. If a suitable resistor connects the anode and cathode an electron

flow can be measured that is generated by the bacteria. In nature, such a microbially initiated

electron flow normally ends by the transfer of electrons to the terminal electron acceptor such as

oxygen, nitrate, sulfate or even poorly soluble minerals (Clauwaert et al. 2007a; Newman and

Kolter 2000; Rabaey et al. 2005b; Zhao et al. 2006a). It has been shown that electron acceptor

couples with a more positive redox potential enable bacteria to oxidize the electron donor more

effectively as it provides a greater amount of Gibbs Free Energy Change (Cord-Ruwisch et al.

1988).

It is conceivable that also the bacteria in MFC profit from a more positive potential of

the external electron acceptor (here the anodic potential), as this increases the energetics (more

negative Gibbs free energy change) of the bacterial reaction. From a biological point of view one

would expect that very negative anodic potential (the anodic potential approaching that of the

electron donor (here acetate)) limits the microbial capacity to pass electrons to the anode.

Further one would expect that at very positive anodic potentials the microbial activity is limited

by factors other than the availability of a suitable electron accepting system and that ―electron

acceptor saturation‖ will be observed, similar to the substrate saturation known for enzymatic

and microbial processes.


- 55 -

MFC offers the possibility of varying the potential of the electron acceptor (here the

anode) without changing the reacting species (e.g. nitrate, oxygen, sulfate), and hence a MFC is

a suitable bio-electrochemical tool to study this fundamental aspect of microbial physiology

(Rabaey et al. 2007). In the literature, the dependency of the microbial metabolic activity on the

anodic potential of a MFC has only been tested with only limited detail and no quantitative

study has been published (Schroder 2007). In a recent study on a benthic bio-electrochemical

system the anodic potential was controlled artificially by a potentiostat and the authors could

show that there was a tendency of increasing microbial activity with increasing anodic potentials

(Finkelstein et al. 2006).

It is the objective of this study to establish a generic understanding of the behavior of

anodophilic bacteria in a MFC when the anodic potential becomes limiting. Similar to the

widely used saturation behavior for soluble species, such knowledge is expected to assist with

evaluating the efficiency of microbial populations in MFC. In order to test the undisturbed

activity, in-situ, we used the anodic potentials as they were generated and changed by the

bacterial biofilm in the MFC. Different approaches were adopted to obtain the relationship

between microbial activity and anodic potential in the MFC. These included: (1) manual step

changes of external resistance as commonly used in many other studies for polarization curve

analysis; (2) constant sweeping of external resistance and (3) open circuit to allow building up of

anodic potential under substrate saturated or limited conditions. The interrelationship among

microbial activity, anodic potential and MFC power output is explored in detail. Further it is

attempted to explain the maximum power output of the fuel cell by the microbial capacity to

discharge electrons to the anode at different anodic potentials.


- 56 -


3.2.1 Microbial Fuel Cell Start-up and On-line Process Monitoring

A dual-compartment microbial fuel cell made of transparent Perspex was used in the

present study. The two compartments were physically separated by a cation selective membrane

(CMI-7000, Membrane International Inc.) with a surface area of 168 cm2. The two

compartments were of equal volume and dimension (316 mL (14 cm x 12 cm x 1.88 cm)).

Conductive granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-6 mm

diameter) was used as both the anode and cathode electrodes, which decreased the internal liquid

volume from 316 to 120 mL. Contacts between the external circuit and the granular graphite

electrodes were made using graphite rods (5 mm diameter).

To start up the MFC experiment, the anode chamber was inoculated with activated

sludge collected from a local wastewater treatment plant (Woodman Point, Perth, WA,

Australia). 10% (v/v) of freshly collected activated sludge with a biomass concentration of about

2.0 g L-1

was mixed with synthetic wastewater. The composition of the synthetic wastewater

was (mg L-1

): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25,

and 1.25 mL L-1


): ethylene-diamine tetra-acetic

acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25,

NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O

0.050. Sterile concentrated yeast extract solution was periodically added (ca. every 3 to 5 days)

to the anolyte (50 mg L-1

final concentration) as growth nutrients.

The MFC was operated in a fed-batch mode with both catholyte and anolyte

continuously re-circulating over the cathode and anode, respectively. Unless otherwise stated,

the recirculation rates were 8.0 L h-1

to minimize mass transfer limitation. Sodium acetate was

used as the sole electron donor for the bacterial culture. For replenishment of electron donors in

the MFC, a designated amount of a concentrated sodium acetate stock solution (1M) was

injected into the anode chamber via a septum-secured injection port just before the inlet of the

anode chamber.

Since the present study is focused on the anodic process of MFC, a strong and stable

cathodic reaction is essential for effective control of the anodic potential. Hence, potassium

ferricyanide (50 - 100 mM), K3Fe(CN)6 (Sigma-Aldrich, Inc., Purity ca. 99 %) was selected as

the terminal electron acceptor in the catholyte despite its non-sustainable nature compared to


- 57 -

oxygen (air)-cathode (Freguia et al. 2007a; Logan et al. 2006). Air-saturated ferricyanide

solution complemented with 100 mM phosphate buffer (pH 7) was continuously re-circulated

through the cathode chamber at a fixed flow rate to reduce cathodic over-potential. Outside of

the MFC the catholyte was circulated through a 2 L aerated glass bottle which contained a gas

diffuser for continuous purging of the solution with water-saturated air at a flow rate of 1 L min-

1. Unless a weaker cathodic process was desired in some experiments, the catholyte was renewed

periodically to maintain a reasonably stable cathodic potential.

Designated flow rates in both chambers were adjusted by using peristaltic pumps. Unless

otherwise stated all experiments were performed at a constant temperature (30oC) and ambient

atmospheric pressure. Control and monitoring of the microbial fuel cell was partially automated.

Potential differences between anode and cathode, Eh in both anolyte and catholyte, and pH of

anolyte were monitored continuously by using LabVIEW™ 7.1 software interfaced with a

National Instrument™ data acquisition card (DAQ). The anodic potential was measured against

a silver/ silver chloride reference electrode (saturated potassium chloride) which was placed

inside the anodic compartment. Accuracies of all the voltage signals detected by the DAQ were

regularly checked by using a high precision digital multimeter (resolution 10 µV). All data were

logged into an Excel spreadsheet with a personal computer (data logging frequency depended on

the signal changes at different experiments).

As virtually no net proton migration across the cation exchange membrane is observed

under normal MFC operating condition (Rozendal et al. 2006a), protons (H+) generated from

microbial oxidation of acetate accumulated, resulting in a pH drop of the anolyte. This pH drop

was found to inhibit microbial activity as reflected by diminished electron flow of the MFC

(data not shown) (Gil et al. 2003; Zhao et al. 2006a). To overcome the problem, the pH of the

anolyte was strictly controlled at 7 ± 0.2 by the automated addition of NaOH (1M) using a

computer-controlled peristaltic pump throughout the whole study. In experiments where acetate-

saturated conditions (2 to 5 mM) needed to be maintained, semi-continuous, automated acetate

dosing was implemented using a computer feedback-controlled peristaltic pump with pre-

defined time interval or the anolyte Eh level as the reference set point in the LabVIEW™

feedback control program.


- 58 -

3.2.2 Calculation and Analysis

3.2.2.1 Determination of Voltage, Current and Power Generation

Fuel cell power output (P) was calculated from the voltage (V) measured across a known

resistor (adjusted by using a variable resistor box) and the current (I). An 8-channel PC parallel

relay board connected to a pre-defined array of resistors was controlled by the computer to allow

computer adjusted external resistance of the MFC. I was calculated according to Ohm‘s Law, I =

V/R, where V is the measured voltage and R (Ω) is the external resistance. Power output was

calculated according to: P=V x I. Since it was impractical to determine the surface area of the

granular graphite electrode used in this study and a significant part of the surface area of the

anode was inaccessible to bacteria, volumetric current density and power density were

determined by normalizing the values to the void volume of the anodic chamber (Rabaey and

Verstraete 2005). Current (in milliamperes, mA) was converted to electrons recovered by using

the following conversions: 1 Coulomb (C) = 1 Amp x 1 second, 1 C = 6.24 x 1018

electrons, and

1 mol = 6.02 x 1023

electrons (96,485 C/mol). Coulombic efficiency was calculated according to

Logan et al. 2006 (Logan et al. 2006).

3.2.2.2 Measuring the Effect of Anodic Potential on Microbial Activity by varying

External Resistance

The relationship between anodic potential and microbial activity in the MFC was

established by varying the external resistance of the MFC to attain steady potential of the anode.

This was achieved with three different approaches:

(1) Manual step changes of external resistance. This technique is commonly used in

many other studies mainly for polarization curve analysis (Logan et al. 2006);

(2) Constant sweeping of external resistance. A computer controllable relay board

equipped with a series of resistors was used to give a desired resistance output. As

such, the external resistance of the MFC could be systematically controlled at a fixed

sweeping rate (unit: Ω min-1

); and

(3) Switching at steady state the external resistance from a low level (e.g. 1 Ω) to open

circuit and observing the decrease of anodic potential. This approach examines the

bacterial capacity to charge up the anode under different conditions (e.g. substrate

saturated or limited conditions).


- 59 -


3.3.1 Characteristics of the MFC

A microbial fuel cell was set up as described above and operated for over 200 days.

During this time the current generated by the bacteria increased steadily. It was established that

the electron flow from anode to cathode was proportional to the rate of acetate oxidation by the

bacteria, and as soon as acetate in the anolyte was depleted, the electron flow in the cell dropped

to close to zero (Figure 3.1). A majority of electrons obtained from acetate oxidation was

retrieved as measured electron flow through the resistor (coulombic efficiency of about 83%)

(Figure 3.2).

Figure 3.1 Effect of external resistance (1 Ω and 1 MΩ) on acetate degradation ( ) and current

production (solid line) in the acetate-fed microbial fuel cell.

3.3.2 Initial Changes of Resistors result in Steady States of Different Microbial Activities

In order to test the effect of the anodic potential on the microbial acetate oxidation rate

the MFC was operated with different resistors under acetate saturated conditions (Figure 3.3).

As expected, a large external resistance (e.g. 160 Ω) did not allow a ready flow of electrons from

the anode to cathode and hence the anode stayed ―charged up‖ by the bacteria resulting in a low

anodic potential of less than -530 mV (vs. Ag/AgCl). For each resistance chosen, a new steady

state anodic potential established. This steady state is the result of the rate of microbial electron

Time (min)

Curr

ent (m

A)

Aceta

te C

oncentra

tion (m

M)

0

20

40

60

80

100

120

140

0 200 400 600 800

0

2

4

6

8

10

0

20

40

60

80

100

120

140

0 200 400 600 800

0

2

4

6

8

10

1 MΩ 1Ω 1 MΩ 1Ω

Acetate addition


- 60 -

delivery towards the anode being identical to the electron flow away from the anode to the

cathode and hence the calculated current can be used as a direct indicator of microbial acetate

oxidation rate (Figure 3.3D).

Figure 3.2 Current ( ) and Coulombic efficiency ( ) as a function of initial acetate

concentration in the acetate-fed microbial fuel cell. (Data were obtained after 16 days of

operation at 1 Ω.)

In general, the use of smaller resistors caused an increase in the anodic potential (Figure

3.3A and B). As expected, the microbial acetate oxidation rate increased with the anodic

potential as this not only provides a higher free energy change of the microbially catalyzed

reaction, but it also stimulates the re-oxidation of any electron shuttles (mediators) that may be

present in the system hence facilitating the electron transfer from bacteria to such mediators.

3.3.3 At a Certain Potential Range the Dependency of Microbial Activity on Anodic

Potential is Less Defined

More careful exploration of the relationship between microbial activity and the anodic

potential revealed that once an anodic potential of about -420mV (vs. Ag/AgCl) was reached, no

clear steady state values of current and potential could be obtained. Small decreases in resistance

did no longer increase the current but instead the anodic potential drifted rapidly to becoming

more positive (Figure 3.3B, at ~110 min).

Acetate Concentration (mM)

Curr

ent (m

A)

0

20

40

60

80

100

120

140

0.0 0.5 1.0 1.5

0

20

40

60

80

100

Ks = 0.36 mM

Cou

lom

bic

Effic

iency (%

) Vmax. = 140 mA


- 61 -

Figure 3.3 Changes in (A) external resistance; (B) anodic potential; (C) current and (D) acetate

oxidation rate in the acetate-saturated microbial fuel cell during a manual step-change of

external resistance experiment. (Note: the values of acetate oxidation rate were

calculated from the current after it had reached steady state. A new resistor was only

chosen after a new equilibrium between anodic potential and current had established.)

-550

-450

-350

-250

-150

-50

0

10

20

30

40

50

60

70

80

0

20

40

60

80

100

120

140

160

180

0

20

40

60

80

20 30 40 50 60 70 80 90

0

2

4

6

8

10

50 100 150 200 250 300

-540-520-500-480-460

0 10 20 30 40 50 60 70 80 90

05

101520

20 25 30 35 40 45 50

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 50 100 150 200 250 300

Time (min)

Aceta

te O

xid

ation R

ate

(mM

∙h-1)

Curr

ent (m

A)

Anodic

Pote

ntial

(mV

vs. A

g/A

gC

l)

Exte

rnal R

esis

tance

(Ohm

)

(B)

(A)

(D)

(C)


- 62 -

The unexpected drift in anodic potential was studied by allowing longer times for the

system to equilibrate (Figure 3.4). After initial steady values of anodic potential and current for

each resistor chosen, the potential started drifting at about -420 mV (vs. Ag/AgCl) when the

external resistance was decreased from 8 to 7 Ω (~8 min). The anodic potential was allowed to

drift resulting in a much higher quasi-steady state of about -165 mV (vs. Ag/AgCl) (~25 min).

Even by using slightly larger resistors the trend could not be reversed. Many repeat experiments

verified this to be a reproducible phenomenon. Overall, reproducible values of microbial activity

for anodic potentials between -350 and -420 mV (vs. Ag/AgCl) could not be obtained.

Figure 3.4 Changes of anodic potential (grey curve) and current (black curve) in the acetate-

saturated microbial fuel cell during downward and upward step-changes of external

resistance. (Dotted arrow shows the drifting of anodic potential after the external

resistance was decreased from 8 to 7 Ω at about 8 min.)

3.3.4 Apparent Maximum of Microbial Activity at Relatively Low Anodic Potentials

In spite of the unstable behavior of current and anodic potential between -350 and -420

mV (vs. Ag/AgCl) a maximum of microbial activity could be observed from a plot of acetate

oxidation rate versus anodic potential (Figure 3.5). Similar trends were observed at different

developmental stages of the anode biofilm, indicating that such phenomenon was reproducible

and independent of biomass concentration and composition in the MFC. The observed small but

marked maximum of apparent microbial activity at a rather low anodic potential between -350

and -420 mV (vs. Ag/AgCl) would indicate that the bacteria preferred a particular anodic

-450

-400

-350

-300

-250

-200

-150

-100

0 10 20 30 40 50 60 70 80

0

10

20

30

40

50

60

70

80

90

100

Time (min)

An

od

ic P

ote

ntia

l

(mV

vs. A

g/A

gC

l)

Cu

rren

t (mA

)

Exte

rnal R

esis

tan

ce

(Oh

m)

30

25

20

15

10

5

0

Current Anodic

Potential

External

Resistance 10Ω

9Ω 8Ω

7Ω 8Ω

10Ω

20Ω

30Ω


- 63 -

potential rather than being most active at the most positive anodic potential. This finding is

surprising for two reasons:

1. An increased anodic potential offers more free energy for the microbial reaction and

would be expected to stimulate microbial activity.

2. A lowering of resistance would normally increase the electron flow of MFC even if the

anodic reaction was not stimulated.

The fact that a lowering of resistance in a MFC caused a lower current is contrary to

general observations reported in the literature (Gil et al. 2003; Logan et al. 2006; Menicucci et al.

2006).

Figure 3.5 Dependency of acetate oxidation rate on the anodic potential of the acetate-saturated

microbial fuel cell. MFC after 5 ( ) and14 ( ) days of operation and after long term

operation with lower currents (with oxygen cathode) ( ).

3.3.5 An Activity Maximum is only obtained when moving from Low to Higher Anodic

Potentials

The phenomenon that at an anodic potential of about -420 mV (vs. Ag/AgCl) the

bacterial activity appeared to have a maximum, was observed for a number of different

conditions as long as incrementally smaller values of resistance were chosen (i.e. sweeping the

potential from negative values towards zero). However, when using incrementally larger rather

0.0

0.5

1.0

1.5

2.0

2.5

3.0

-550 -450 -350 -250 -150 -50 50


Aceta

te O

xid

ation R

ate

(m

M h

-1)


- 64 -

than smaller resistors to document the current-anodic potential relationship, the previously

observed current maximum at around -420 mV (vs. Ag/AgCl) was not found (Figure 3.6).

Specific sweeps of resistance changes both upwards and downwards (sweeping rate of 0.15 Ω

min-1

) established clearly that the observed peak in current was not due to a true maximum in

microbial activity or fuel cell performance but due to a ―hysteresis effect‖ caused only by

increasing the anodic potential. By forming an average of the measurements taken from different

experiments using both increasing and decreasing resistor changes, a reasonably reliable

relationship between bacterial activity and anodic potential could be obtained (Figure 3.6).

Figure 3.6 MFC current ( = by decreasing external resistance; = by increasing external

resistance; = average current) as a function of anodic potential of the acetate-saturated

microbial fuel cell in a constant external resistance sweeping experiment. (Sweeping rate

~0.15 Ω min-1

, ranged from 1 – 78 ohm.)

3.3.6 Open Circuit Drop in Anodic Potential

Another way of investigating the bacterial capacity to charge up the anode is by

recording the build-up of anodic potential at open circuit with a known bacterial acetate

oxidation rate (and hence microbial electron flow generation) (Figure 3.7). After switching the

external resistance from 1 ohm to open circuit, the anodic potential dropped to about -420 mV

(vs. Ag/AgCl) (Figure 3.7A). Then, exactly in the area of the instability of anodic potential and

current as described above, the anodic potential stayed almost constant at about -420 mV (vs.

Ag/AgCl) for about 18 minutes (Figure 3.7A) before droping at a higher rate down to its final


Cu

rre

nt

(mA

)

0

10

20

30

40

50

60

70

80

-500 -450 -400 -350 -300 -250 -200 -150 -100 -50


- 65 -

level at about -480 mV (vs. Ag/AgCl). The same trend could be observed for bacteria that were

starved of acetate. The anode biofilm in Figure 3.7B was starved at high anodic potential for

over 24 hours to oxidize not only all acetate but also any internal storage polymers. This leaves

maintenance respiration as the only electron donating reaction for current production. This

tendency of a decelerating drop in anodic potential between -415 and -425 mV (vs. Ag/AgCl)

suggests one of two possibilities:

1. The bacterial electron release was now ―buffered‖ by the presence of an oxidized

electron mediator (capacitance). In this case, instead of decreasing the potential further

the mediator is being reduced (charged up).

2. Alternatively the slower drop in potential could be due to the bacteria becoming less

active.

Figure 3.7 Profile of anodic potential (black curve) and anolyte Eh (grey curve) of a microbial

fuel cell under (A) acetated-saturated and (B) acetated-depleted conditions during open

circuit. (Dashed line indicates the anodic potential at which quasi-steady states were

attained for a short period before dropping further to the lowest steady level.)

3.3.7 Detailed Interpretation of the Dependency of Microbial Activity on Anodic Potential

Figure 3.8 illustrates a plot of microbial activity and MFC power output as a function of

terminal electron acceptor availability (anodic potential). The curve was characterized by the

following values:

1. At anodic potentials lower than about -420 mV (vs. Ag/AgCl) the microbial activity was

limited by the anodic potential or by the availability of sufficient oxidized electron

mediator that is used for the transfer of electrons to the anode.

-500

-400

-300

-200

-100

0

100

0 10 20 30 40 50 60 70 80 90 0 60 120 180 240 300 360 420 480Time (min)

Po

ten

tia

l (m

V v

s. A

g/A

gC

l)

1Ω 1Ω

Anode

Open Circuit

(A) Substrate Saturated (B) Substrate Depleted

Anolyte

Anode

Anolyte

Open Circuit

Time (min)


- 66 -

2. There was no significant microbial activity at anodic potentials less than -480 mV (vs.

Ag/AgCl). This is most likely caused by the reducing capacity of the electron donor

acetate with standard redox potential of -480 mV (vs. Ag/AgCl). The fact that some

current was observed at anodic potentials lower than that could be explained by the

product and substrate concentrations affecting the actual Eh relative to the standard Eh for

standard conditions. Moreover, bacteria may have the capacity to generate more powerful

reducing agents in the process of assimilation.

3. There was a Plateau of microbial activity for anodic potentials between -200 and about -

420 mV (vs. Ag/AgCl) representing excess availability (saturation) of electron acceptors

(oxidized mediators or anode). The microbial metabolism was not limited by the capacity

of the anode to serve as electron sink but by other metabolic bottlenecks such as acetate

uptake, TCA cycle or similar.

4. In biochemical terms a half saturation constant for anodic potential could be estimated. In

analogy to the Michaelis-Menten half saturation constant (here termed kAP value) it would

represent the anodic potential (or the concentration of oxidized mediator) at which the

metabolic rate (here measured as current) reaches half its maximum rate. This kAP value

was about -455 mV (vs. Ag/AgCl) for our acetate driven MFC and was independent of

the oxidation capacity of the cathodic half cell.

5. There was a critical anodic potential at about -420 mV (vs. Ag/AgCl) (APcrit., Figure 3.8)

beyond which further increased anodic potentials no longer stimulated bacterial activity

significantly, showing that now the biochemical electron flow was saturated by the

availability of the terminal electron acceptor (in this case the oxidized mediators for the

anode or the anode itself).

The significance of the above described critical anodic potential (APcrit.) is that it

characterizes the power output of the MFC (Figure 3.8). At anodic potentials below the APcrit.

the power output of the fuel cell increases with lower resistors providing a higher anodic

potential and enabling a greater microbial electron flow. At anodic potentials higher than APcrit.

the bacteria cannot generate significantly more electron flow as they are saturated by the

availability of suitable electron acceptors. Here, a further increase in anodic potential does not

yield more power because the current no longer increases substantially while the voltage of the

MFC diminishes due to increasing anodic potential (Figure 3.8). This suggests that the maximal


- 67 -

power output of a MFC is close to the point where the bacteria reach saturation of their electron

accepting mechanism by the anode.

Figure 3.8 Percentages of maximal microbial activity ( = manual R step-change; = constant

R sweeping; solid curve) and MFC power output ( = manual R step-change; =

constant R sweeping; dashed curve) as a function of anodic potential of an acetate-

saturated microbial fuel cell. (kAP = half saturation constant for anodic potential, ~ -455

mV vs. Ag/AgCl; APcrit. = critical anodic potential, ~ -420 mV vs. Ag/AgCl); Notes:

constant R sweeping rate was ~0.15 Ω min-1

; catholyte Eh ranged from 225 to 232, and

260 to 300 mV (vs. AgAgCl) for constant R sweeping and manual R step change

experiments, respectively.

The observation that a half saturation value can be determined and is in analogy to the

half saturation constants for the concentration of soluble substrates seems quite interesting and is

not published yet. It is expected to be useful for modeling and optimization of microbial fuel

cells and the relative comparison of different microbial consortia at the anode. However, as it is

known for kinetic values obtained for activated sludge (e.g. the kO2, the half saturation

concentration for the terminal electron acceptor oxygen) the APcrit. and kAP values are not

expected to be discrete values for all other MFC systems. However the general behavior is

widely used for modeling and process optimization. The discrete value would depend on the

thickness of the biofilm, the mass transfer limitation in the anodic chamber, reactor

configuration, the species composition present, similar to the kO2 value for activated sludge

bacteria.

0

20

40

60

80

100

-500 -400 -300 -200 -100 0

A Pcrit.

K AP


Perc

enta

ge o

f M

axim

al A

ctivity

and P

ow

er

(%)


- 68 -

3.3.8 Redox Capacitance as an Explanation of Apparent Current Maximum

At this stage we assume that the peak in current observed during stepwise increases of

the anodic potential (but not during stepwise decreases) was not due to increased bacterial

activity but to reducing power that had been stored at the previous, lower potentials. This

tendency was also indicated by the Eh trends in the anodic half cell (Figure 3.7). It is not likely

that the well known phenomenon of bacterial storage (e.g. as PHB) formation (under anaerobic

conditions) and its oxidation under aerobic condition is responsible for this reproducible

observation because: a) the time intervals were only short, b) the same phenomenon could be

observed with acetate starving and acetate saturated cells. The fact that when using non-

equilibrium conditions by moving stepwise down with the resistors gave a different curve than

by moving stepwise up is in analogy to cyclic voltametry. Even though the ―sweeping rate‖ used

in our experiments is very low in comparison the resultant lines suggest that capacitance has

build up in the system.

However, it goes beyond the scope of this study to investigate the source (mediators in

the bulk solution, inside the biofilm or as part of the electrode) capacitance of this system.

However it can be stated that significant charge storage is observed. If the capacitance was due

to microbial shuttling compounds, then an approximate mid-point potential (or standard

potential) between -350 and -450 mV (vs. Ag/AgCl) would be indicated (Figure 3.6, 3.7).

3.3.9 Implication of Findings on the Design and Operation of MFCs

The fact that the microbial activity in MFC obeys the same kinetic type for the anode

than described for soluble species (Michaelis-Menten kinetics) is expected to be helpful in

evaluating different biofilms, bacteria species or environmental conditions. Similar to the

saturation kinetics of microbial metabolism of soluble species, it is expected that some biofilm

populations have a high affinity for the anode while others display a high maximum activity.

With the new insight about the relationship between microbial activity and anode

potential, an optimal operating condition of MFC could be defined by controlling or maintaining

the anodic potential at the APcrit. level at which the maximal bacterial activity and power output

of the MFC could be attained. Further, knowing the actual affinity for the anode and maximum

microbial activity in a MFC is expected to help with diagnosing fuel cell malfunctions.

- 69 -

4 A New Approach for in-situ Cyclic Voltammetry of a

Microbial Fuel Cell Biofilm without using a

Potentiostat

(This chapter has been published in Bioelectrochemistry, (2009) 74: 227-231)

Chapter Summary

Electrochemically active bacteria in a microbial fuel cell (MFC) usually exist as a

biofilm attached to an electrode surface. Conventional cyclic voltammetry using potentiostat is

considered as a powerful and reliable method to study electrochemical behavior of MFC biofilm.

In this chapter, a new approach to evaluate redox behavior of an electro-active MFC biofilm

without using a potentiostat was proposed and validated. Analogous to a conventional cyclic

voltammetry study, the biofilm-electrode potential was controlled by computer-feedback

controlling the external resistance of an operating MFC. In this way, the MFC can still operate

as a ―fuel cell‖ without being ―interrupted‖ by an external device (i.e. potentiostat) that normally

does not belong to the system.

Relationship between current and biofilm-electrode potential was obtained and showed

agreement with a potentiostat-controlled method under similar experimental conditions. The

method could be added to our technical repertoire for analysis of bacterial mediator involved in

the exocellular electron transfer of a MFC-biofilm, and it could potentially serve as a practical

process monitoring method for MFC operation. The application of computer-control

components should be further explored to facilitate control, diagnosis as well as optimization of

MFC processes.

Chapter 4: A New Approach for CV analysis without using a Potentiostat

- 70 -

4.1 Introduction

In light of the increasing demand of renewable and sustainable energy, microbial fuel

cell (MFC) technology becomes one of the most rapidly developing environmental

biotechnologies (Logan and Regan 2006b; Lovley 2006b; Rabaey and Verstraete 2005). Yet, the

technology is still in its infancy and needs further research to improve understanding and

performance of the process. MFC is a bacteria-catalyzed electrochemical device, in which a

bacterial biofilm on the anode transfers electrons from the oxidation of an organic substrate to

the anode enabling a current flow from anode to cathode. The presence of a suitable resistive

load in the external circuit enables the production of electrical power.

Arguably the most interesting phenomenon in MFC electron flow is the transfer of

electrons from the bacteria to the anode surface. Although different mechanisms have been

proposed, the most widely accepted view is that electron mediators are produced by the bacteria

for this transfer. Bacterial mediators (also termed electron shuttles or electron carriers) have also

been found in other microbial reactions involving an insoluble solid electron acceptor such as

ferric iron or elemental sulfur (Lovley et al. 2004; Stams et al. 2006). After the bacteria have

reduced an oxidized mediator as their terminal electron acceptor, it will then ―shuttle‖ the

electrons to the electron accepting anode where it becomes reoxidized, and becomes available to

the bacteria again (Rabaey et al. 2004; Rabaey et al. 2007; Schroder 2007).

The redox property of the bacterial mediators can be seen as critical for the effective

operation of the MFC. In the literature, redox properties of mediators were commonly studied by

using cyclic voltammetry (CV) method. This method is considered as a standard technique for

studying redox behavior of solutions (Bard and Faulkner 2001; Logan et al. 2006; Rabaey et al.

2005a; Rabaey et al. 2004; Scholz 2002; Sund et al. 2007).

In CV, the potential of a working electrode is scanned forward and backward at a

predefined rate (commonly ranges from 5 to 100 mV·s-1

). The presence of any mediators is

detected from the fact that a current is produced during the cyclic oxidation and reduction of the

mediator, provided that the mediator can accept (/ donate) electrons from (/ to) the working

electrode. The mid potential of the mediator can then be obtained from the current-voltage plot

(i.e. cyclic voltammogram) (Bard and Faulkner 2001; Marken et al. 2002).

Using conventional CV (i.e. using potentiostat to control a half cell potential) to evaluate

electrochemical behavior of an electrode-associated biofilm has several advantages. For


- 71 -

examples: (i) the target electrode processes (e.g. biofilm-anode) can be studied separate from the

opposing electrode (i.e. cathode) reaction and separate from other side effects such as ion

transfer processes across the separating membrane; (ii) the potential can be varied in a wide

range, even exceeding the potential range of a chemical cathode.

However, in conventional CV the MFC has to be disconnected from the external power

user (i.e. external resistance) in order to connect with a potentiostat. From a practical viewpoint,

this may lead to an operation down time of the process during which the MFC can no longer

generate power. In this chapter, a new method of controlling the MFC external resistance to

perform cyclic voltammetric analysis of an anode biofilm is proposed. In this way, the MFC can

still operate as a ―fuel cell‖ without being ―interrupted‖ by an external device (i.e. potentiostat)

that normally does not belong to the system. Further, the optimal external resistance of a MFC at

which the MFC delivers the highest power may also be defined or maintained by using the

proposed method.

It is the aim of this study to develop a practical approach of performing CV analysis for

MFC process without using a potentiostat. This method will supplement the well established

conventional CV method of evaluating redox properties of an electrode-associated biofilm. It is

hypothesized that CV of a MFC biofilm could be done by feedback-controlling the external

resistance of an electricity producing MFC without the need of a potentiostat. This means no

additional equipment is needed other than the computer-controlled MFC itself. The MFC was

set-up and operated as described in Chapter 3.


- 72 -


4.2.1 Underlying Principle of the Proposed Method

Under close-circuit operation of an active MFC (i.e. electrons are allowed to flow readily

from the biofilm anode to the cathode), there is a continuous reduction and reoxidation of the

mediator involved in the electron transfer from the bacteria and the anode. The reduction

reaction can be realized by using a larger external resistance, resulting in an accumulation of

reduced mediator and hence charging up the biofilm-anode (i.e. decreased electrode potential).

While the oxidation reaction is facilitated by using a smaller external resistance, resulting in

faster anode discharge (i.e. increased electrode potential).

Similar to the theory of traditional CV our proposed method aims to recognize mediators

by their capacity to donate or accept electrons when the electrode potential is continuously

shifted (scanned). However, instead of using an external reducing power source (i.e.

potentiostat), our proposed method relies solely on the bacteria to reduce their self-produced

mediators using electrons obtained from their substrate oxidation. In this case the electrode

serves only as the final electron acceptor. This method can selectively detect redox species

involved exclusively in the bacterial electron transfer chain, and therefore it has the potential to

offer more specific information compared to traditional CV in evaluating redox properties of a

MFC biofilm.

4.2.2 Microbial Fuel Cell

A two-chamber MFC made of transparent Perspex was used in the present study. Figure

4.1 shows the schematic of the reactor and its peripheral settings. The two chambers were of

equal volume and dimension (316 mL (14 cm 12 cm 1.88 cm)). They were physically

separated by a cation-selective membrane (CMI-7000, Membrane International Inc.) with a

surface area of 168 cm2. Conductive graphite granules (El Carb 100, Graphite Sales, Inc.,

Chagrin Falls, OH, U.S.A., granules 2-6 mm diameter) were used as both the anode and cathode

electrodes, which decreased the internal liquid volume from 316 to 120 mL. This internal void

anolyte volume was used to calculate the volumetric current and power densities of the MFC.

The biofilm in the anode chamber was originated from an activated sludge collected

from a local wastewater treatment plant. The growth medium (anolyte) composed of (mg L-1

):


- 73 -

NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL

L-1


): ethylene-diamine tetra-acetic acid (EDTA)

15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O

0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. The

medium was complemented with 50 mM phosphate buffer to maintain a constant pH of 7. Yeast

extract solution was periodically added (ca. every 3 to 5 days) to the medium (50 mg L-1

final

concentration) as bacterial growth supplement. Air-saturated solution of potassium ferricyanide

(100 mM), K3Fe(CN)6 (Sigma-Aldrich, Inc., purity ca. 99%) complemented with 100 mM

phosphate buffer was used as the catholyte. Detail operation and monitoring of the MFC during

the initial start-up period were described in our previous paper (Cheng et al. 2008).

Figure 4.1 Schematic of the computer-feedback controlled microbial fuel cell used in the

present study.

4.2.3 Cyclic Voltammetry by Feedback-Controlling the External Resistance

The biofilm-electrode potential was controlled by selecting appropriate external

resistance of the MFC. In initial experiments, the external resistance of the MFC was

continuously changed in a linear fashion by using an eight channel computer controlled relay

board (Ocean Controls, Victoria, Australia; www.oceancontrols.com.au, Photo 4.1) (Figure

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Stirrer

1M NaOH

Acetate Feed

Moisten

Air




Catholyte Eh

Ag/AgCl

Reference

Electrode

Biofilm-Anode

(Granular Graphite)

Counter Electrode

(Granular Graphite)


e- e-

Recirculation

Pump

MF

C V

olta

ge

Bio

film

-Ele

ctr

ode

Po

ten

tial

Catholyte

Graphite Rod

( Diameter =

5mm )

Out In

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Stirrer

1M NaOH

Acetate Feed

Moisten

Air




Catholyte Eh

Ag/AgCl

Reference

Electrode

Biofilm-Anode

(Granular Graphite)

Counter Electrode

(Granular Graphite)


e- e-

Recirculation

Pump

MF

C V

olta

ge

Bio

film

-Ele

ctr

ode

Po

ten

tial

Catholyte

Graphite Rod

( Diameter =

5mm )

Out In

http://www.oceancontrols.com.au/


- 74 -

4.2A). The relay was directly controlled by the parallel port output (8 bits) from the computer.

The slow rate linear sweeping of external resistance (0.69 min-1

) had reproducibly yielded the

so-called polarization and power density curves which are commonly used to assess MFC

performance (Logan et al. 2006) (Figure 4.2B). Furthermore, this constant resistance sweeping

technique was validated in our previous study to establish the relationship between the current

and the anodic potential (Figure 4.2C), from which the affinity of an electrochemically active

biofilm for the electrode potential could be determined (Chapter 3, Cheng et al. 2008).

Photo 4.1 Computer controllable relay board served as the variable resistor of the MFC.

In practice, the external resistance can be controlled in a number of ways, including the

use of an electronically adjustable resistor. In our experiments the switching of 8 individual

resistors via the relay board allowed up to 256 different resistance values, which was sufficient

for adequate control of the biofilm-electrode potential. This improved way of changing the

resistors allowed a relatively linear scan of potential over time (Figure 4.3A and B). This was

accomplished as follows: a computer program was designed using LabVIEW™ programming

software (Version 7.1) to allow the operation of the MFC at a constant electrode (here anode)

potential setpoint. To maintain the electrode potential setpoint the external resistance was

controlled (see above) by a simple feedback loop in the program, which enabled not only the

operation to a desired potential setpoint but also the programmed stepwise shift of the potential.

This could allow a linear potential change over time (here 0.1 mV·s-1

), as it is known in CV (i.e.

upward and downward scans represent shifts of potential toward a less and a more negative

values, respectively) (Figure 4.3B).


- 75 -

Figure 4.2 Constant sweeping of the external resistance of the computer-feedback controlled

acetate-saturated microbial fuel cell (0.69 min-1

). (A) Resistance vs. time; (B)

Polarization and power density curves; Note: ↑ or ↓ R represents increasing or decreasing

sweep of external resistance, respectively. (C) Current vs. electrode potential.

0

20

40

60

80

-600 -500 -400 -300 -200 -100 0


Curr

ent

(mA

)

0

100

200

300

0 5 10 15Time (h)

Resis

tance (

Ohm

) .

0

100

200

300

400

500

600

700

0 200 400 600


)

Voltage (

mV

)

0

50

100

150

200

250

300

Pow

er D

ensity

(W m

-3)

Voltage: ↑ R

Voltage: ↓ RPower: ↑ R

Power: ↓ R

E/ mV (vs. Ag,AgCl (3M KCl))

Voltage/ m

V

Curr

ent/

mA

Power D

ensity

/ W∙m

-3

A.

B.

C.

Time/ h

Resis

tance/ O

hm

Current Density/ A∙m-3


- 76 -

It should be noted that these experiments were carried out with a fully developed biofilm

that was saturated with organic substrate supply (here acetate, about 10 mM) which served as a

relatively constant source of reducing power in place of a potentiostat. The ultra low scan rate of

0.1 mV·s-1

was selected to allow sufficient time for the biofilm (microbially-reduced mediator)

to interact with the oxidizing electrode. As a result, the method enabled the detection of

mediators that were not only located within the boundary layer of the electrode-solution

interface, but also mediators far away from the electrode (i.e. confined within the biofilm or

inside the bacterial cells). Using this method with scan rates that are normally used in traditional

CV, no significant peaks/ valleys were detected (data not shown).

4.2.4 Cyclic Voltammetry using a Three-electrode Potentiostat

In the present study, the new method was validated by comparing with results obtained

from a potentiostat-controlled method. In this Control experiment, the potential of the same

biofilm-electrode was regulated by using a three-electrode scanning potentiostat (Model no.362,

EG&G, Princeton Applied Research, Instruments Pty. Ltd.). The working, counter and reference

electrodes of the potentiostat were connected to the anode, cathode and the silver-silver chloride

reference electrode of the acetate-saturated MFC, respectively. The electrode potential was

changed at a similar scan rate of 0.1 mV·s-1

to allow reasonable comparison with the new

method. To assure quality of the results, all experimental data (obtained with both the new

method and the Control method) were validated for reproducibility by running for at least three

repetitive cycles in each experimental testing.


- 77 -


4.3.1 Feedback Controlling the External Resistance of a MFC enables Cyclic

Voltammetry of MFC Biofilm

The current obtained during the scans of biofilm-electrode potential was recorded and

showed characteristic and reproducible patterns (Figure 4.3C). In theory, decreasing electrode

potentials will thermodynamically impede the electron flow from the bacteria to the electrode

and hence resulting in lower currents. However, depending on whether the potential was

gradually increased or decreased, different current lines were obtained (Figure 4.3B and C). This

is indicative of charging and discharging of bio-electrochemically active compounds. To

visualize more clearly the redox behavior of the electrochemically active biofilm the current was

plotted as a function of electrode potential (i.e. cyclic voltammogram).

Figure 4.3 Time profiles of (a) external resistance; (b) electrode potential; and (c) current during

the computer-feedback control of the external resistance of an acetate-saturated microbial

fuel cell. Electrode potential scan rate was 0.1 mV·s-1

.

-450

-350

-250

-150

Ele

ctr

od

e P

ote

ntia

l

(mV

vs. A

g/A

gC

l)

0

20

40

60

80

100

0 1 2 3 4

Cu

rre

nt (m

A)

0

20

40

60

80

Re

sis

tan

ce

(O

hm

)

Time (h)

A.

B.

C.


- 78 -

To test the hypothesis of the present study, the new method was verified by comparing

with results obtained from a potentiostat-controlled method. Figure 4.4 indicates that both

computer-feedback- and potentiostat-controlled experiments revealed similar results: the upward

shift of electrode potential from about -380 to -200 mV vs Ag/AgCl gave significantly higher

currents compared to currents obtained from the downward shift in the same potential range.

While at potentials outside this range, the resultant currents recorded from both upward and

downward scans were virtually identical. Many repeat experiments verified this to be a

reproducible phenomenon. These reproducible equilibrium currents (i.e. bacterial activity) were

independent on the direction of potential scans and indicated that bio-electrochemically active

species were not active (charging/ discharging) at these potentials (Figure 4.4).

Figure 4.4 Comparison between the current-potential plots (cyclic voltammograms) obtained by

(a) computer-feedback controlling the external resistance and (b) by using a potentiostat

in an acetate-saturated microbial fuel cell. Scan rates of the electrode potential in both

methods were 0.1 mV∙s-1

; dotted arrows indicate the direction of the potential scan; the

difference between the currents obtained from the two methods was due to the different

times of the two experiments.

The drop in current when the potential decreased from about -200 to -330 mV vs

Ag/AgCl suggests one of two possibilities:

0

20

40

60

-450 -350 -250 -150


Cu

rre

nt

(mA

)

Feedback Controlling of Resistor

Potentiostat-Controlled

35

40

45

-450 -350 -250 -150

E/ mV (Ag,AgCl (3M KCl))

Cu

rre

nt/

mA


- 79 -

(1) Bacteria became less active at lowered electrode potentials;

(2) The electrons released by the bacteria were accepted and stored by a ―pool‖ of alternative

electron acceptor (i.e. oxidized mediators) instead of the electrode (capacitance from

mediator).

Since a further decrease in potential from -330 to -380 vs Ag/AgCl resulted in increasing

currents again, the first possibility seems unlikely (Figure 4.4). The observed current ―valley‖ in

both experiments was likely caused by the redox reactions of the mediator. This is supported by

the fact that during upwards shifts of potentials there was no ―valley‖. From the cyclic

voltammograms obtained in these modified CV experiments (Figure 4.4), one could obtain the

mid potential of the mediator from the potential at which the upward and downward curves are

most widely apart (i.e. highest capacitance). In our acetate fed MFC biofilm, such potential

value was about -330 mV vs Ag/AgCl.

It has to be noted that as MFC is a bio-electrochemical system, the anode associated

biofilm can grow and develop over time. Hence, over hours of operation during the CV

experiments, it is expected that the current output obtained at even the same electrode potentials

would change. This explains the difference between the current-potential curves in Figure 4.4.

Nevertheless, our results suggested that both the proposed and the conventional CV methods

indicated a same redox behavior of the same biofilm.

However, it is speculative to characterize the nature of a redox active species involved in

the anodic electron transfer in a MFC (i.e. whether it is purely membrane associated such as

cytochrom c or nanowire (i.e. direct electron transfer) or soluble (i.e. indirect electron transfer)

solely based on results obtained from both the proposed and the conventional CV techniques.

According to earlier findings (Bond and Lovley 2005; Rabaey et al. 2007), bacterial soluble

mediators appear to be confined mainly within the biofilm rather than in the bulk medium. This

property was due to an electrostatic interaction between the soluble mediator and the anode

(Rabaey et al. 2007). Therefore, CV of a biofilm-anode cannot be used as a stand-alone method

to ascertain the physiochemical nature of the mediator involved in the electron transfer process.

Overall, the new method has demonstrated a good controllability of the biofilm electrode

potential, and it showed a similar capacity to the conventional potentiostat-controlled method in

evaluating in situ electrochemical properties of a MFC biofilm. Further, provided that the

electron donor is not limiting in the system, evaluation of the electro-active biofilm in a MFC


- 80 -

can be conducted without the need of a potentiostat. To our knowledge, this is the first report in

the literature to control the biofilm-electrode potential by feedback controlling the external

resistance of a MFC in a way similar to a CV analysis. Redox information about the mediator(s)

involved in the bacterial electron transfer as well as the affinity of the electro-active biofilm on

the electrode potential (i.e. values of critical electrode potential and half-saturation electrode

potential as defined in (Chapter 3, Cheng et al. 2008; Marcus et al. 2007) could be obtained with

minimal disturbance of the bioprocess.

4.3.2 Limitations and Implications of the New Method

The established technique has two major limitations that are needed to be addressed in

future studies.

1. It requires the presence of bacterial substrate. In MFC, as the electrons flow exclusively in

one direction starting from the substrate bacteria bacterial mediator(s) anode

external resistor cathode terminal electron acceptor, sufficient substrate (electron

donor) must be provided to the biofilm during the analysis. While in the absence of

substrate, there is no reducing power to reduce the involved mediator(s) and hence the

electrode. Perhaps, such criterion could be further explored for investigating specific

electron transfer processes with different substrates.

2. It requires an efficient cathodic reaction. The cathodic reduction process should not be

limiting over the electrode potential range in which bioelectrocatalysis of the bacterial

mediators is taking place. In the present study, the use of ferricyanide solution enabled a

stable and strong cathodic reduction, allowing the effective control of anodic potential by

using the proposed method. However, it is expected that with the continuous improvement

of MFC cathodic process, such limitation could be alleviated.

Apart from the application as discussed above, the established technique may also be

useful to gain other important insights for MFC process control. For instance, selection of

external resistor is considered as critical for optimization of MFC process. Maximal and

sustainable power output can only be achieved when the MFC is running with an appropriate

external resistance. While most studies reported thus far have only arbitrarily selected the

external resistance to operate their MFCs, and very often the power outputs of their systems

were reported by using an inappropriate external resistor (Menicucci et al. 2006). By feedback


- 81 -

controlling the external resistance of the MFC using either current/ power generation as the

reference parameter, we could in theory operate the MFC at its optimal current/ power

production point (see supplementary figure in Appendix 3). Overall, it is expected that more

sophisticated and intelligent application of computer control components will be emerged in the

future to deal with optimization and diagnosis of such dynamic bio-electrochemical process.

- 82 -

5 An Anodophilic Biofilm prefers a Low Electrode

Potential for Optimal Anodic Electron Transfer: A

Voltammetric Study

Chapter Summary

The knowledge of electron transfer between anodophilic biofilms and their associated

electrodes is important for both the fundamental understanding and optimization of BES. In this

Chapter, a series of potentiostatic experiments were conducted to explore in details the electron

transfer behavior of an active anodophilic biofilm in a BES. The results suggest that using ultra

slow potential scan rates of 0.01mV s-1

(between -450 and -150 mV) an actively current

producing bacteria revealed peaks of current that could not be detected with traditional scan

rates or with starving bacteria. It appeared as if the bacterial electron transfer to the electrode

had an optimum potential at around -230 mV vs. Ag/AgCl. At another potential of -310 mV vs.

Ag/AgCl the presence of a redox species (or redox system) was indicated by the fact that a peak

during upwards (from -ve towards +ve) potential scanning was not found during downwards

scanning. The draining of the electrolyte and replacing by fresh solution did not affect the

electrochemical characteristics of the anodic biofilm, demonstrating that the electrochemically

active species were predominately present in the biofilm at the electrode rather than in the bulk

medium.

The fact that the presence of the redox peaks was dependent on the presence of the

bacterial electron donor (here acetate) suggested that the reduction of the mediators (redox links)

was of biochemical and not purely electrochemical nature. The fact that the complexity of the

bioelectrochemical behavior of an electrochemically active biofilm could only be revealed by

using very low scanning rates indicates that the redox process in the biofilm-anode system was

slow compared to abiotic mediator-electrode systems. Overall, the use of low scan rates for CV

analysis may be an improved way to characterize the electrochemical behavior of an anodophilic

biofilm in BES.

Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential

- 83 -

5.1 Introduction

As mentioned in Chapter 1 and 3, the potential of the anode determines the amount of

Gibb free energy conserved by the anodophilic bacteria in a BES (denoted as ΔGo‘

bacteria in

Figure 1.5, page 14). From a thermodynamic viewpoint, it is expected that the anodophilic

bacteria could conserve more energy at a higher anodic potential and hence a more positive

anodic potential is expected to encourage establishment of electrochemically active bacteria on

the electrode surface. However, in order to maximize electricity generation in MFC or to

minimize the extra energy input in MEC (both are denoted as ΔGo‘

Elec in Figure 1.5), a minimal

amount of ΔGo‘

bacteria is preferred as it keeps the anode at a lower redox state and hence the

anode can be more reducing. Further, if the bacterial energy gain becomes too large (e.g. at a

very positive anodic potential), the electrical energy output will be diminished and electrons

from the substrate may be diverted toward an assimilation metabolism, leading to excessive

biomass yield but reduced electron recovery (Pham et al. 2009). This may lead to a question of

whether an anodophilic biofilm would prefer or adjust their metabolism to a certain anodic

potential to meet the antagonistic requirements between bioenergetics and thermodynamics of a

BES? In the literature, there are evidences suggesting that anodophilic biofilms exhibit

saturation behavior with electrode potentials (Chapter 3, Cheng et al. 2008; Torres et al. 2008a;

Torres et al. 2007). Yet, no clear evidence is available in the literature to elucidate the existence

of a preferred electrode potential for optimal anodic electron transfer. Moreover, the underlying

process of anodic electron transfer from electrochemically active microorganisms to their

associated electrode (bioanode) is highly complex (Fricke et al. 2008; Pham et al. 2009;

Schroder 2007).

Anodophilic bacteria in BES commonly exist as a biofilm attached to the electrode, the

diffusion or mass transport of soluble mediators (if any) either toward or away from the

electrode surface is limited as the biofilm may block the direct electrochemical contact between

the mediators in the bulk and the electrode. Therefore, to better qualify or quantify the redox

behavior of an electrochemically active biofilm using cyclic voltammetry, sufficient time is

required to allow the redox active species to response to the electrode potential.

In Chapter 4, a new approach of performing voltammetric analysis in a BES without

using an external electrochemical device (a potentiostat) has been described. However, as

mentioned in Section 4.3.2 this new method adheres to two inflexibilities: 1) the anodic biofilm

must be saturated with the electron donor substrates; and 2) a powerful cathodic reaction must


- 84 -

be provided. While the use of a potentiostat to investigate the electron transfer of an

electrochemically active biofilm can avoid disturbance and artifact due to the second electrode

(i.e. cathode, if anode is the electrode of interest) or the internal resistance of the set-up.

In this chapter, a potentiostatic mini-scale bioelectrochemical reactor was designed to

study in details the biofilm-anode interaction. The method used here tries to evaluate redox

activity not only in the boundary layer of the electrode but also mediators that are located deeper

inside the biofilm. This is accomplished by using ―ultra‖ low scan rates (0.01 mV s-1

) which

enables a more pronounce influence of the electrode potential on redox active species (mediators)

that are not readily accessible by the electrode (i.e. far away from the electrode, confined within

the biofilm or inside the biomass).


- 85 -


5.2.1 Anodophilic Biofilm and Growth Medium

The electrochemically active anodophilic biofilm used in this study was established from

an acetate-enriched MFC that had been operated for over 200 days. Detail operation of the MFC

was described in Chapter 3. In brief, the biofilm was derived from an activated sludge inoculum

obtained from a local domestic wastewater treatment plant. During MFC operation, the anodic

potential lied mostly in a range between -300 and -450 mV vs. Ag/AgCl. At day 216, about two-

thirds of the biofilm coated anode granules were harvested from the MFC and were immediately

immersed in a growth medium to preserve the bacterial viability before they were packed into

the bio-electrochemical cell used in this study. Excess biofilm covered granules were stored in

the medium at 4 oC for later use. The growth medium consisted of (mg L

-1): NH4Cl 125,

NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL L-1

of trace

element solution, which contained (g L-1

): ethylene-diamine tetra-acetic acid (EDTA) 15,

ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O

0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. The

medium was complemented with 50 mM phosphate buffer to maintain a constant pH of 7.

5.2.2 Construction and Operation of Bio-electrochemical Cell

Figure 5.1 illustrates the schematic diagram of the three-electrode bio-electrochemical

cell used in this study. Similar to a conventional two-chamber MFC, it consisted of a working

and a counter chamber. The biofilm covered electrode was placed inside the working chamber,

which could accommodate a volume of about five cubic centimeters of the graphite granules.

Electrical contact between the graphite granules and the potentiostat was secured by using a

graphite rod current collector (5 mm diameter), which was embodied and pierced through a

rubber cover lid of the working chamber. A platinum sheet and a silver-silver chloride reference

electrode (3 M KCl) served as the counter and reference electrodes, respectively.

To minimize mass transfer limitation inside the working chamber, the working medium

was recirculated over the granular graphite at a flow rate of about 10 mL·min-1

by using a

peristaltic pump (Cole-Parmer Masterflex, model 7519-00). The recirculation tubing was

impermeable to oxygen (Tygon® , lumen diameter 3 mm) and was wrapped with aluminium foil

to avoid growth of photo-heterotrophic bacteria. The counter electrolyte was 100 mM phosphate


- 86 -

buffer solution (pH 7). It was continuously mixed by using a magnetic stirrer bar and was

regularly renewed throughout the potentiostatic study.

Figure 5.1 (A) Schematic diagram and (B) photo of the mini-scale bio-electrochemical system used for

the potentiostatic experiments. (1) Graphite rod as current collector (Five millimeters diameter);

(2) silver-silver chloride reference electrode; (3) Biofilm coated graphite granules; (4) Separator

membrane (Nafion 117 proton exchange membrane); (5) Platinum foil; (6) 100 mM phosphate

buffer (pH 7); (7) Magnetic stirrer bar.

Since ultra slow potential scans of the working electrode were used in the study, it is

essential to avoid electrochemical contact between any electrochemically active species of the

V

i

LabVIEW™

Potentiostat

Counter Working Reference

InflowOutflow

(1 )(2 )

(4 )

(5 )

(6 )

(7 )

(3 )

V

i

LabVIEW™

Potentiostat

Counter Working Reference

InflowOutflow

(1 )(2 )

(4 )

(5 )

(6 )

(7 )

(3 )

(A)

(B)


- 87 -

biofilm system and the counter electrode. Hence, instead of placing the three electrodes into a

common liquid medium, a proton exchange membrane (Nafion 117, Fuel cell store™) was used

to separate the biofilm covered working electrode as well as its associated growth medium from

the reference and the counter electrodes. This configuration was designed to simulate a MFC

operation with the use of a potentiostat to maintain a precise biofilm-electrode potential.

All experiments were conducted under well-controlled electrochemical conditions. Prior

to each experiment, the platinum counter electrode was pre-treated by heating in a flame to a

red-heat condition followed by quenching in deionized water. The biofilm electrode (i.e.

working electrode) was polarized against the reference electrode at desired potential value by

using a computer-programmable potentiostat (manufactured and quality assured by Murdoch

University Electronic Workshop).

A computer program written with LabVIEW™ (version 7.1 National Instrument) was

used to control and manipulate signals of the potentiostat. An analog input/ output data

acquisition board (DAQ) (Labjack™, UE9, 12-bits of resolution) was used as an analog-digital

converter to interface between the computer and the potentiostat. Voltage and current from the

potentiostat were recorded at fixed time intervals and the data were logged into an Excel

spreadsheet with a personal computer.

Accuracies of all the voltage signals from the DAQ were verified by using a high

precision digital multimeter (resolution 10 µV). Quality of the electrochemical data generated by

the system was satisfactorily assured by running cyclic voltammograms of a standard

ferricyanide solution (various concentrations complemented with 50 mM phosphate buffer, pH 7)

(data not shown). To avoid abiotic evolution of hydrogen or oxygen at the electrode during the

potentiostatic studies, the potential of the working electrode was only controlled in a narrow

range (anywhere between 0 and -0.6 V vs Ag/AgCl). This potential range was selected in order

to simulate the real operating anodic condition of MFC.


- 88 -


To investigate electrochemically active species in solution cyclic voltammetry is usually

the method of choice. Most commonly, this method is based on using clean and defined

electrodes to oxidize and reduce dissolved mediators present in the vicinity of the working

electrode. However, in the case of bioelectrochemical systems such as microbial fuel cells

electron shuttling species are present with the biofilm growing on the electrode surface. In order

to evaluate potential electrochemically active species in the anodic compartment of the microbial

fuel cell described, cycling potentials were applied directly by using the biofilm attached

electrode as the working electrode in a bio-electrochemical cell (Figure 5.1).

Figure 5.2 Cyclic voltammograms of biofilm covered graphite granules from the anode of an

established MFC under electron donor (acetate) saturated and starved conditions. Scan

rate was 0.1 mV·s-1

.

5.3.1 Only Actively Metabolizing Biofilm Exhibited Electrochemical Activity in Cyclic

Voltammetry

When just using the biofilm covered graphite granules as the working electrode without

the supply of acetate, and employing CV at a scan rate of 0.1 mV·s-1

no significant

electrochemically active species could be detected (Figure 5.2). However, when the biofilm was

supplied with saturating concentrations of its electron donor (10 mM acetate) there was

background anodic current observed at most potentials in both scanning directions, suggesting

that the working electrode would serve as an anode for the biofilm (Figure 5.2). This

-0.10

0.00

0.10

0.20

0.30

-0.45 -0.35 -0.25 -0.15

Potential (V vs Ag/AgCl)

Cu

rre

nt (m

A) Acetate Saturated

Acetate Starved


- 89 -

phenomenon is expected and has also been described by Fricke et al. (2008). Such anodic

current is not caused by the voltage applied by the potentiostat but by the bacteria, provided that

the electrode potential was held sufficiently high (> -380 mV vs. Ag/AgCl) to enable bacterial

transfer of electrons from acetate to the electrode.

Figure 5.3 Effect of acetate depletion on the current profile during repetitive cyclic voltametry

of the active, MFC enriched anodophilic biofilm covered graphite granules with a scan

rate of 0.1 mV·s-1

. Asterisks indicate the characteristic current peaks observed.

The peaks in the cyclic voltammogram obtained under substrate saturation (Figure 5.2)

suggest the presence of an electrochemically active species in the biofilm with potentials ranging

from -320 to -220 mV vs Ag/AgCl. To verify this effect, a limiting amount of the electron donor

of the biofilm (acetate) was added to the cell and repetitive cycles of CV at 0.1 mV·s-1

were run.

Results showed that initially the peaks of the CV were as described above and could be shown

as current peaks over time (Figure 5.3). As the acetate concentration diminished, the current

measured during the cycles changed by:

the decrease in overall current as the bacteria no longer had an electron donor.

the disappearance of the redox peaks obtained (depicted as asterisks in Figure 5.3).

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 2 4 6 8 10 12 14

Time (hour)

Cu

rre

nt (m

A)

-0.45

-0.4

-0.35

-0.3

-0.25

-0.2

-0.15

Po

ten

tial (V

vs. A

g/A

gC

l)

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 2 4 6 8 10 12 14

Time (hour)

Cu

rre

nt (m

A)

-0.45

-0.4

-0.35

-0.3

-0.25

-0.2

-0.15

Po

ten

tial (V

vs. A

g/A

gC

l)

Acetate sufficient Acetate diminishing

Cycle: 1st 2

nd 3

rd 4

th 5

th 6

th 7

th


- 90 -

After addition of acetate to the depleted cell, the previous characteristic peaks and current

levels resumed. To test whether the peaks in the cyclic voltammograms were caused by potential

mediators present in the bulk liquid, the bulk liquid was drained and replaced by fresh medium.

This had no effect on the shape of the cyclic voltammogram (data not shown), supporting the

findings obtained by Fricke et al. (2008) in a Geobacter bioelectrochemical cell showing that the

redox active species are predominately located inside the biofilm or as part of the bacterial cells.

5.3.2 Redox Mediators of the Active MFC Biofilm were Located Far Away from the

Electrode Surface

When applying CV on the acetate saturated and hence actively current generating cell, a

reproducible current-potential pattern occurred (Figure 5.4). However, at scan rates commonly

used for detecting mediators in MFC systems (i.e. ranging from 5 to 50 mV·s-1

) (He et al. 2005;

Logan et al. 2006; Qiao et al. 2008; Rabaey et al. 2004), no particular peaks of electrochemically

active species could be detected (Figure 5.4). This could be due to the fact that the

electrochemically active species in the biofilm may not be in the immediate vicinity of the

electrode (i.e. the boundary layer) as it is known for CV with clean and defined electrodes.

Figure 5.4 Cyclic voltammograms of an active, acetate saturated biofilm covered graphite

granules (anode) of an established MFC with (A) scan rates commonly used for

conventional electrochemical study of mediator in MFCs and (B) very low scan rates.

It was considered that the scan rate should be further reduced to allow the applied

potentials to affect mediators further away from the electrode potentially situated within the

-3

-2

-1

0

1

2

3

-0.55 -0.45 -0.35 -0.25 -0.15 -0.05 0.05

Potential (V vs Ag/AgCl)

Curr

ent

(mA

)

5 mV/s10 mV/s25 mV/s50 mV/s

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

-0.45 -0.4 -0.35 -0.3 -0.25 -0.2 -0.15

Potential (V vs. Ag/AgCl)

Curr

ent

(mA

)

0.2 mV/s0.4 mV/s0.8 mV/s

(A) (B)


- 91 -

biofilm or even inside the bacterial cells. Therefore, the scan rate was further reduced to an

extremely low rate of 0.01 mV·s-1

with the same biofilm. Although, this ultra low scan rate was

rarely reported in the literature, this could offer a longer equilibration time for the biofilm to

response to the effect of electrode potential.

Clearly, such an extremely slow scan revealed the presence of two (instead of one)

reproducible and distinct peaks at about -310 and -240 mV vs Ag/AgCl, respectively (Figure 5).

The two peaks obtained during the increasing potential scan (i.e. upper curve) signified either (a)

the discharge of mediators that the bacteria had previously reduced at lower potentials; or (b)

increased capacity of bacteria to produce electricity at higher rates at these higher potentials.

Figure 5.5 Ultra slow cyclic voltammetric scan (0.01 mV·s-1

) of an actively acetate

metabolizing MFC biofilm.

5.3.3 Detail Interpretation of the Anodophilic Properties of the MFC Biofilm in CV

In order to determine which of the above statement is more likely to explain the peaks

obtained, a more detailed scan was recorded around the voltage of the observed peaks. Upon

more close investigation of the electrochemical behavior of the biofilm by running scans around

the two key peaks at -310 mV and -240 mV vs Ag/AgCl, a key difference could be detected

between the two peaks (Figure 5.6).

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

-0.55 -0.45 -0.35 -0.25 -0.15 -0.05 0.05

Potential (mV vs. Ag/AgCl)

Curr

ent

(mA

)

Peak 2 Peak 1


- 92 -

Figure 5.6 Cyclic voltammetry with a narrow potential range covered (A) around the first peak

at -310 mV vs. Ag/AgCl and (B) around the second peak at -230 mV vs. Ag/AgCl. Scan

rate was 0.01 mV·s-1

.

The peak at -310 mV vs. Ag/AgCl was only observed during upward scans while

downward scans showed no such peak (Figure 5.6A). Instead the downwards scan showed a

gradual decrease in current caused by the deceleration of bacterial electron transfer rate. This

peak is more in line with the traditional view that a maximum in electron flow is caused by the

discharging of previously charged electron mediators as also observed by other authors (Fricke

et al. 2008).

0.00

0.05

0.10

0.15

0.20

0.25

0.30

-0.33 -0.32 -0.31 -0.30 -0.29 -0.28 -0.27


Cu

rre

nt (m

A)

0.10

0.11

0.12

0.13

0.14

0.15

0.16

0.17

0.18

-0.26 -0.25 -0.24 -0.23 -0.22 -0.21


Cu

rre

nt (m

A)

(A) 1st Peak:

(B) 2nd

Peak:


- 93 -

In contrast, the peak at -230 mV vs. Ag/AgCl was visible in both scanning directions

(Figure 5.6B). This phenomenon is not in line with typical CV results. It cannot be explained by

the charging and discharging of electrochemically active compounds by the electrode. This

suggests that this second peak is not due to the redox behavior of mediator but that it is caused

by a maximum in bacterial activity at the potential of about -230 mV, independent on whether

previously higher or lower electrode potentials were applied.

5.3.4 Step-Change of Electrode Potential Signified the Existence of an Optimal

Electrode Potential for Current Production from the Biofilm

If the peak at -230 mV vs. Ag/AgCl observed above is truly due to the microbes showing

increased metabolic activity at this potential, then this maximum should also appear under

steady state conditions at a fixed potential. In order to verify whether there is such an optimum

electrode potential at which the anodic current shows a maximum, the MFC was run at fixed

potentials for extended periods of 1 h, before step-changing to a different potential, which was

again kept for 1 h (Figure 5.7).

Figure 5.7 Microbially generated current flow with acetate as electron donor at different

electrode potentials in a constant potential step-change experiment (1 h per step).

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0 2 4 6 8 10 12 14

-0.34

-0.32

-0.30

-0.28

-0.26

-0.24

-0.22

-0.20

-0.18

Current

Potential

Time (h)

Curr

ent (m

A)

Pote

ntia

l (V v

s A

g/A

gC

l)

Peak 4 Peak 1 Peak 3

Peak 2


- 94 -

The resulting current profile confirmed that the biofilm always gave a higher current

between -220 and -240 mV vs. Ag/AgCl than at either –180 to -200 mV vs. Ag/AgCl or at < -

260 mV vs. Ag/AgCl. Such current maximum remained when step changing the potentials either

upwards or downwards (Figure 5.8).

Figure 5.8 Steady state current as a function of electrode potential (data from Figure 5.7). Note:

the average current was calculated from data obtained in the last 30 min of the respective

potential.

Figure 5.9 Microbially generated current flow with acetate as electron donor at three applied

electrode potentials across the second peak at -230 mV vs. Ag/AgCl under steady state

conditions (1 h constant potential).

0.00

0.05

0.10

0.15

0.20

-0.35 -0.30 -0.25 -0.20

Potential (mV vs. Ag/AgCl)

Cu

rre

nt (m

A)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0 1 2 3 4 5 6 7

Time (hour)

Cu

rre

nt

(mA

)

-0.255

-0.245

-0.235

-0.225

-0.215

-0.205

-0.195

-0.185

Po

ten

tial (V

vs. A

g/A

gC

l)

Current

Potential


- 95 -

In contrast, the peak at -310 mV vs. Ag/AgCl could only be seen during stepwise

increasing potentials (Figure 7, 0 – 3 h), while decreasing potentials (Figure 7, 11 – 14 h) did not

show such a peak. Again the results suggest that the two peaks observed are of a different nature.

The dependency of the current on the potential changes shown above was very

reproducible in the cell tested. In a final attempt to test whether the MFC biofilm truly produced

a higher current at a lower and hence for the bacteria less attractive electrode potential, again the

results showed that indeed at anodic potentials between - 225 and -250 mV vs. Ag/AgCl the

bacteria transferred electrons to the anode faster than at -190 mV vs. Ag/AgCl (Figure 5.9).

5.3.5 Implication of Findings

This chapter suggests that with cyclic voltammetry the electrochemical activity of a

MFC biofilm could only be detected when the biofilm was actively metabolizing and generating

an electron flow, but not when the same bacteria were starved of their electron donor. This is not

in line with the idea that the mediators indicated by the peak in the cyclic voltammogram

(Figure 5.2) were oxidized and reduced by the electrode directly. The results can be explained

by the concept that the mediators detected in this ―biochemical CV‖ were unable to be reduced

by the electrode and were strictly dependent on active bacterial metabolism. While it is clear that

this species actively participates in the electron transfer from bacteria to the electrode, it is not

likely that it is the terminal electron carrier as it would by definition be able to be re-oxidised by

the graphite electrode itself.

The possibility of an optimum anodic potential at which the bacteria can cause a

maximum of current could be significant during optimization of MFCs. In the literature there

were indications of such an optimum to exist (Aelterman et al. 2008; Cheng et al. 2008; Torres

et al. 2007), however it has not been identified or quantified as the current study does. In terms

of microbial activity and of thermodynamics it would be expected that the highest potential of

the electrode would enable the fastest electron flow from the bacteria to the electrode. The

reason for the presence of a particular electrode potential of about -230 mV vs. Ag/AgCl beyond

which the microbial current generation decreases is unclear. However, the measured electron

flow (current) only measures the rate at which acetate is being oxidized (dissimilation). Under

growing conditions a proportion of the acetate taken up is not dissimilated but assimilated which

does not result in a flow of electrons. Accordingly it is perceivable that if the acetate intake rate


- 96 -

becomes limiting and at increasing potentials the generation of ATP from respiration becomes

more readily possible that a lowering of electron flow signifies an increase in the metabolic

quotient (i.e. respiration per unit biomass) of the cells.

On the other hand, it can be envisaged that the established biofilm (which was

predominately cultivated at a low operating anodic potential in the previous experiments, < -300

mV vs. Ag/AgCl) may have discontinued to synthesize internal electron carriers that are

responsible for shuttling electrons to a high potential electron acceptor (at which the biofilm had

rarely exposed to), but rather the bacteria had switched their metabolism in synthesizing the

electron carrier(s) that are more thermodynamically compatible with (or had a higher affinity to)

a specific, more negative anodic potential. In nature, this situation would be similar to

facultative bacteria that can stop expressing electron carriers required for oxygen reduction

under anaerobic condition (Van Spanning et al. 1995). For example, a facultative nitrate-

reducing bacterium Paracoccus denitrificans was found to denitrify (nitrate) even at an oxygen

level of 100% air saturation (Sears et al. 1997). Here, the question of whether the anodophilic

biofilm can adapt their electron shuttling machinery in favor of the redox potential of their

associated electrode still remains open. Yet, this question could be valuable not only to BES

application and also to fundamental microbiology.

Further researches are warranted to elucidate in detail the electron transfer mechanism

(preferably at a protein-level) of the observed dependency between the electrode-living biofilm

and its surrounding redox environment.

- 97 -

6 Alternating Bio-Catalysis of Anodic and Cathodic

Reactions alleviates pH Limitation in a BES

— An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction —

(This chapter has been published in Environ. Sci. Technol. (2010) Vol.44(1)518-525)

Chapter Summary

As stated in Chapter 1, the poor cathodic oxygen reduction and the detrimental build-up

of electrolyte pH are fundamental hurdles in sustainable microbial fuel cell (MFC) development.

This chapter describes and tests a concept that can help overcoming both of these limitations, by

inverting the polarity of the MFC repeatedly, allowing anodic and cathodic reaction to occur

alternately in the same half-cell and hence neutralizing its respective pH effects.

For simplicity, we studied polarity inversion exclusively in one half cell, maintaining its

potential at -300 mV (vs. Ag/AgCl) by a potentiostat. The alternating supply of acetate and

dissolved oxygen to the biofilm resulted in the tested half cell to turn repeatedly from anode to

cathode and vice versa. This repeated inversion of current direction avoided the detrimental

drifting of the electrolyte pH. Control runs without current inversion ceased to produce current,

due to anode acidification.

The presence of the anodophilic biofilm survived the intermittent oxygen exposure and

could measurably facilitate the cathodic reaction by reducing the apparent oxygen overpotential.

It enabled the cathodic oxygen reduction at about -150 mV (vs. Ag/AgCl) compared to -300 mV

(vs. Ag/AgCl) for the same electrode material (granular graphite) without biofilm. Provided a

suitable cathodic potential was chosen, the presence of ―anodophilic bacteria‖ at the cathode

could enable a 5-fold increase of power output.

Overall, the ability of an electrochemically active biofilm to catalyze both substrate

oxidation and cathodic oxygen reduction in a single bioelectrochemical system has been

documented. This property could be useful to alleviate both the cathodic oxygen reduction and

the detrimental drifting of electrolyte pH in a MFC system. Further research is warranted to

explore the application of such bidirectional microbial catalytic property for sustainable MFC

processes.

Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction

- 98 -

6.1 Introduction

Microbial fuel cells (MFCs) represent a bio-electrochemical process that can directly

convert chemical energy stored in organic substances into electrical energy. In theory, any

biodegradable organic substance can potentially serve as the ―fuel‖ for this process. Therefore,

MFCs are seen as a promising waste-to-energy environmental technology. Yet, there are many

bottlenecks that hinder their development as a sustainable energy conversion process. Apart

from other upscale technical hurdles such as material costs and reactor configurations, the poor

cathodic oxygen reduction and the establishment of a pH gradient between anodic and cathodic

regions are currently perceived as the two major fundamental hurdles in MFC development

(Logan 2008; Rabaey et al. 2008; Rozendal et al. 2006a; Zhao et al. 2006a).

Continuous operation of MFC causes acidification at the anode due to the accumulation

of protons liberated from the microbial oxidation of organic compounds (eq. 6-1), whereas

alkalinization at the cathode is observed due to continuous consumption of protons by the

oxygen reduction reaction (eq. 6-2) (Zhao et al. 2006a). Given that oxygen accepts four

electrons, then four moles of protons are required for each mole of oxygen molecule used.

CH3COO- + 4 H2O 2 HCO3

- + 9 H

+ + 8 e

- (eq. 6-1)

2 O2 + 8 H+ + 8 e

- 4 H2O (eq. 6-2)

Ideally, these protons are replenished from the anodic oxidation of organic substrate. However,

ionic species such as sodium, potassium, ammonium, magnesium, and calcium will

preferentially migrate across a cation membrane as their concentrations are commonly much

higher (Rozendal et al., 2006). Consequently, the continuous consumption of protons for the

oxygen reduction reaction results in a net increase of catholyte pH, while the anode will become

acidified due to accumulation of protons. These lead to a so-called membrane pH gradient

overpotential or pH splitting limitation, which puts an electrochemical/ thermodynamic

constrain on the MFC performance (Harnisch et al. 2008; Rabaey et al. 2008; Rozendal et al.

2008a).

Harnisch et al. (2008) have recently highlighted the proton migration mechanism and

problems associated with the use of proton exchange-, cation exchange-, anion exchange- and


- 99 -

bipolar membranes in MFC systems. They concluded that none of these membrane separators

could circumvent the pH splitting problem, which leads to diminished bacterial activity and poor

cathodic oxygen reduction. To rectify the problem, active pH control such as acid/ base dosing

or addition of chemical buffer (e.g. phosphate buffer or bicarbonate) are required to sustain

steady current production. However, these approaches are not practical as they represent a

significant cost and energy input (Harnisch and Schroder 2009; Pham et al. 2009; Rozendal et al.

2008a).

Another key hurdle of MFC technology is the poor cathodic reaction when oxygen is

used as the terminal electron acceptor. Although oxygen is considered to be the most sustainable

terminal electron acceptor for MFC, the poor kinetics of oxygen reduction (i.e. high cathodic

overpotential) under conditions prevailing in most MFC configurations (i.e. neutral electrolyte

pH, low electrolyte ionic strength and ambient temperature) diminishes the performance of

oxygen cathode (Freguia et al. 2007b; Yu et al. 2007). The use of platinum as catalyst is not seen

as a sustainable solution because of costs and environmental impact in its production (Freguia et

al. 2007b; He and Angenent 2006). Recent studies show that bacteria can play a role in the

cathodic oxygen reduction (Clauwaert et al. 2007a; Freguia et al. 2007c; Rabaey et al. 2008;

Rozendal et al. 2008b) and promise to become an alternative to the conventional approach.

However, their potential is not fully explored.

Recently, it has been shown that the catalytic activity of an anodophilic biofilm is not

compromised by the presence of oxygen (Biffinger et al. 2008; Malki et al. 2008). Whether an

anodophilic biofilm could also assist oxygen reduction at a cathode has not been investigated so

far. If a biofilm could catalyze both anodic substrate oxidation and cathodic oxygen reduction,

an alternating current would be obtained when exchanging substrate and oxygen supply from

one half cell to the other. Since the anodic substrate oxidation is a proton liberating process

while the cathodic oxygen reduction is a proton consuming process, such inversion of MFC

polarity may potentially alleviate the pH limitation during MFC operation.

The current study aims at addressing the problem of pH gradient as well as the slow

cathodic oxygen reduction by intermittently reversing the polarity of the MFC. The specific

research questions are:

1) Can the detrimental build-up of electrolyte pH be avoided?

2) Can anodophilic bacteria catalyze the cathodic oxygen reduction?


- 100 -

As a proof-of-concept study, a potentiostatically controlled two-chamber MFC was used

to answer the above two questions.


- 101 -


6.2.1 Electrochemically Active Biofilm and Growth Medium

The electrochemically active biofilm used in this study was enriched in the anodic

chamber of an acetate-fed MFC during continuous operation for over 200 days (Chapter 3,

Cheng et al. 2008). The initial inoculum before that time was activated sludge. During MFC

operation, the anodic potential was between -200 and -450 mV vs. Ag/AgCl. The growth

medium consisted of (mg L-1


FeSO4·7H2O 6.25, and 1.25 mL L-1


): ethylene-

diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99,

CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014,

and NaWO4·2H2O 0.050. Yeast extract solution was periodically added (every 4 days) to the

medium (50 mg L-1

final concentration) as bacterial growth supplement. This medium was used

as the electrolyte in the working chamber throughout the entire study. A similar medium but

complemented with 50 mM potassium phosphate buffer (pH 7) was used as the counter

electrolyte. Both the working and counter electrolytes were renewed regularly.

6.2.2 Construction and Monitoring of the Bioelectrochemical System

Figure 6.1 illustrates a schematic diagram of the electrochemical cell and its peripheral

settings. A two chamber electrochemical cell with identical electrode material was used as

described in Chapter 2-4. The two chambers were of equal volume and dimension (316 mL (14

cm 12 cm 1.88 cm)). They were physically separated by a cation-selective membrane (CMI-

7000, Membrane International Inc.) with a surface area of 168 cm2. Both chambers were filled

with identical conductive granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-

6 mm diameter). This reduced the void volume of the working chamber from 316 to 160 mL. A

total volume of 350 mL of the electrolyte was re-circulated through the working half cell.

The electrode covered with the anodophilic biofilm described above was defined as the

working electrode, while the other served as the counter electrode. The working electrode was

polarized against a silver-silver chloride reference electrode (saturated KCl) at defined potential

values by using a computer-controllable potentiostat (manufactured and quality assured by

Murdoch University Electronic Workshop). The reference electrode was placed inside the


- 102 -

working chamber, with a distance between the reference electrode and the anode of less than one

cm. This is in contrast to the previous chapters (Chapter 4 & 5, Cheng et al. 2008) where the

reference electrode was 15 cm away from the working electrode, resulting in an average of about

40 mV lower potential readings. This difference has been taken into account in Figure 6.7A. All

electrode potentials (mV) described in this article refer to values against Ag/AgCl reference

electrode (3 M KCl, RE-5B, Bioanalytical Systems, Inc., West Lafayette, Indiana, USA; ca.

+197 mV vs. standard hydrogen electrode (Bard and Faulkner 2001)).

Figure 6.1 Schematic diagram of the potentiostat-controlled two-chamber bioelectrochemical

system used in this study. RE = reference electrode; WE = working electrode; CE =

counter electrode.

A computer program LabVIEW™ (version 7.1 National Instrument) was developed to

continuously control and monitor the bioprocess. An analog input/ output data acquisition card

(Labjack™, U12) was used as an analog-digital converter to interface between the computer and

the potentiostat. Voltage signals from all the monitoring devices were recorded at fixed time

intervals via the LabVIEW software interfaced with a National InstrumentTM

data acquisition

card (DAQ). All data were logged into an Excel spreadsheet. Accuracies (less than 0.1 mV) of

Moisten Air

Stirrer

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Substrate Feed

Moisten

Air

pH Eh pH

Ag/AgCl

Reference

Electrode

Biofilm-Working

Electrode

(Granular Graphite)

Counter Electrode

(Granular Graphite)


Recirculation

Pump

Counter

Electrolyte

Graphite Rod

( Diameter = 5mm)

Out In

Aquarium

Air Pump

DI Water

PC-ControllablePotentiostat

RE WE CE

One Way

Exhaust

DO

DO

Moisten Air

StirrerStirrer

DAQ Board

Computer

Control/

Monitoring by

LabVIEW™

Substrate Feed

Moisten

Air

pH Eh pH

Ag/AgCl

Reference

Electrode

Biofilm-Working

Electrode

(Granular Graphite)

Counter Electrode

(Granular Graphite)


Recirculation

Pump

Counter

Electrolyte

Graphite Rod

( Diameter = 5mm)

Out In

Aquarium

Air Pump

DI Water

PC-ControllablePotentiostat

RE WE CE

One Way

Exhaust

DO

DO


- 103 -

all the voltage signals from the DAQ were verified periodically by using a high precision digital

multimeter (resolution 10 µV).

Oxidation-reduction potential (Eh), pH and dissolved oxygen (DO) concentration

(Mettler Toledo, InPro6800; 4100 PA D/O Transmitter) were continuously monitored at various

locations as specified in Figure 1. Acetate concentration was determined by using a gas

chromatography (GC) as described (Cheng et al. 2008).

6.2.3 Operation of the Bioelectrochemical Process

6.2.3.1 General Operation

To minimize mass transfer limitation, both electrolytes were continuously agitated by

pumping (8 L h-1

) through a separate recirculation loop. The supply of acetate and oxygen

(moist atmospheric air) to the working chamber was computer controlled. A peristaltic pump

(Cole-Parmer, Masterflex®

C/L®

) and an aquarium air pump were used to deliver the acetate

feed (1 M) and air via the inflow of the working chamber at a defined flow rate, respectively. All

experiments were conducted at 30 ± 2 oC.

6.2.3.2 Alternating Supply of Acetate and Oxygen to the Anodophilic Biofilm

The proposed process consists of two phases: (I) an anoxic substrate supply phase and (II)

an aerobic oxygen supply phase (Figure 6.2). During both phases the same electrolyte was used.

In Phase I acetate was added as a spike of 0.5 mmole (resulting in 1.4 mM to simulate

the concentration of a wastewater) to the working chamber, in which the electrochemically

active biofilm attached to the working electrode oxidized the acetate using the working electrode

as the electron acceptor. Acetate was added as a spike rather than continuously fed to allow its

depletion prior to the next phase where the electrolyte was oxygenated, as the presence of both

acetate and oxygen will enable their direct use by the bacteria resulting in less current output of

the cell (Oh et al. 2009).

After 2.5 hours oxygen was introduced into the working electrolyte commencing Phase

II. At this stage acetate had been shown to be consumed as determined by gas chromatography.


- 104 -

The lack of acetate towards the end of Phase I was also indicated by a characteristic drop in

current from typically above 200 mA to close to zero (Figure 6.4A). The oxygen supply to the

electrolyte caused the desired change in polarity, as the potentiostatically controlled working

electrode now became the cathode and resulting in negative current. The DO was online

monitored to ensure that it did not become limiting. Oxygen uptake rate (OUR) measurements

could be derived from online monitoring of the DO at the inflow and outflow as described below.

After 0.5 hours of operating with continuously aerated working electrolyte, the polarity

of the cell was reversed again to its original polarity by stopping oxygen supply and injecting 1.4

mM of acetate to the same working electrolyte.

Figure 6.2 Conceptual diagram of intermittent supply of electron donor and acceptor to an

anodophilic biofilm in a two-chamber potentiostatic microbial fuel cell.

Although the potential of the working electrode was maintained constant (-300 mV for

most experiments and +200 mV for Figure 6.5), the working electrode would serve as either an

anode or a cathode, depending on whether an anodic oxidation of electron donor (acetate) or a

cathodic reduction of acceptor (oxygen) was dominated at the working electrode. Overall, this

approach enabled the biofilm to neutralize the accumulated acidity or alkalinity in the electrolyte

on a sustainable basis. Hence, an active pH control of the electrolyte (e.g. acid/base dosing or

provision of extra pH buffer) was not needed.

Time

Action

Substrate Addition

Aeration Off

Action

Aeration On (O2 Supply)

Oxidation

CH3COO- +4H2O 2HCO3

- +

9H+ + 8e

-

Reduction

2O2 + 8H+ + 4e

- 4H2O

PHASE I (Anaerobic) PHASE II (Aerobic)

Effects

Anodic Current

pH Drop

Effects

Cathodic Current

pH Rise


- 105 -

6.2.3.3 Cathodic Electron Balances

The continuous monitoring of the DO concentrations (DO, mmol O2 L-1

) in the inflow

(DOin) and outflow (DOout) of the working chamber allowed On-line estimation of the OUR at

the working electrode. By considering the flow rate (FR, L h-1

) of the working electrolyte, the

OUR in Phase II could be deduced according to eq. 6-3.

FR

DODO inout )()hO (mmol OUR 1-

2

(eq. 6-3)

4OUR)h e (mmolreduction Ofor rate flowElectron -1-

2 (eq. 6-4)

The OUR is further converted into its corresponding electron flow rate according to eq. 6.4. By

integrating this electron flow rate and the actual cathodic electron flow rate (eq. 6-5) over time,

allowed establishing of a charge balance of the cathodic reaction (i.e. cathodic coulombic

efficiency). Where I is the current (mA) and F is Faraday‘s constant (96,485 C mol e-1

).

F

I3600)h e (mmolcurrent cathodic as rate flowElectron 1-- (eq. 6-5)

6.2.4 Catalytic Effect of the Anodophilic Biofilm on Cathodic Oxygen Reduction

A separate experiment was conducted to establish the relationship between the cathodic

current and electrode potentials. Prior to the experiment, the working chamber was flushed with

fresh electrolyte in order to remove any residual acetate in the system. Thereafter, the electrolyte

was continuously aerated to obtain DO of >3 mg L-1

. Steady state cathodic currents were

recorded and plotted against the electrode potentials ranging from -50 to -400 mV. Abiotic

control experiments were performed using abiotic plain graphite granules as the working

electrode, while all other operating parameters were identical to those in the biotic settings. At a

given potential, the differences in reaction velocity (here current) caused by the biofilm were

considered as catalysis.


- 106 -


6.3.1 Acidification of Electrolyte diminishes Anodic Current Production

At an electrode potential of -300 mV, the addition of acetate could immediately initiate

an anodic electron flow, yielding an anodic current peak of about 240 mA (1500 A·m-3

void

working chamber volume) (Figure 6.4A). Upon acetate depletion (data not shown), the current

started to decline to the original background level.

Figure 6.3 Effect of acetate spikes on current in a potentiostatic microbial fuel cell: (A) with

electrolyte pH controlled at 7; and (B) without pH control of electrolyte. Arrows indicate

acetate additions (450 moles per spike); biofilm-electrode was polarized at -300 mV;

dissolved oxygen in electrolyte was zero.

0

20

40

60

80

100

120

140

Cu

rre

nt (m

A)

4

5

6

7

pH

0

20

40

60

80

100

0 2 4 6 8 10 12 14Time (h)

Cu

rre

nt (m

A)

4

5

6

7

pH

CurrentpH

A.

B.


- 107 -

Figure 6.4 Effect of intermittent supply of acetate and oxygen on (A) current; (B) pH; (C)

electrolyte Eh; (D) inflow/outflow dissolved oxygen concentrations and (E) cathodic

electron flow rate calculated from OUR and current in the biofilm-contained working

chamber of a potentiostatic microbial fuel cell. Note: black arrows indicate acetate

addition; working electrode potential was poised at -300 mV throughout the experiment.

Since the microbial oxidation of acetate is a proton liberating process (see eq. 6-1), the

electrolyte pH decreased over time. Unless pH control is used, this decrease in pH eventually

6.5

6.6

6.7

6.8

6.9

7.0

7.1

pH

-150

-50

50

150

250

Cu

rre

nt (

mA

)

0

0.1

0.2

0.3

DO

(m

mo

l O2/

L)

.

Inflow

Outflow

0

2

4

6

8

0 2 4 6 8 10 12 14 16

Ca

tho

dic

Ele

ctr

on

Flo

w

Ca

lcu

late

d

(mm

ole

e- / h

) .

Time (h)

From OUR

From Cathodic Current

-300

-200

-100

0

100

200

300

Eh

(m

V v

s. A

g/A

gC

l)

6.5

6.6

6.7

6.8

6.9

7.0

7.1

pH

-150

-50

50

150

250

Cu

rre

nt (

mA

)

0

0.1

0.2

0.3

DO

(m

mo

l O2/

L)

.

Inflow

Outflow

0

2

4

6

8

0 2 4 6 8 10 12 14 16

Ca

tho

dic

Ele

ctr

on

Flo

w

Ca

lcu

late

d

(mm

ole

e- / h

) .

Time (h)

From OUR


-300

-200

-100

0

100

200

300

Eh

(m

V v

s. A

g/A

gC

l)

A E R A T I O N

A E R A T I O N

A E R A T I O N

A E R A T I O N

A E R A T I O N

Acetate

Addition


- 108 -

impedes the microbial oxidation of acetate. Figure 6.3A shows that when the pH was strictly

controlled at a set-point of 7, the anodophilic biofilm catalyzed acetate oxidation over prolonged

times. At a poised potential of -300 mV maximum anodic currents of 118 6.7 mA and

coulombic recoveries of 83.3 2.8 % were obtained (Figure 6.3A). However, without the

control of electrolyte pH the continuous build-up of acidity in the electrolyte hampered the

current producing activity of the anodophilic biofilm (Figure 6.3B). At pH 5.2, acetate

addition did not result in a clear current response. A similar pH dependent behavior of the MFC

biofilm has also been reported by Clauwaert et al. (2007b).

6.3.2 Alternating Supply of Acetate and Oxygen to a Biofilm limits Anodic Acidification

It is the intent of this chapter to test whether the detrimental pH drop in the anode can be

prevented by turning the anode intermittently into cathode (inversion of polarity), by merely

replacing the supply of acetate (electron donor) with oxygen (electron acceptor). Rather than

swapping over the oxygen supply from one electrode against the acetate supply of the other (it is

difficult to interpret results as both electrodes suddenly change their behavior), only the working

electrode was studied when supplied alternately with acetate and oxygen. This was done by

maintaining its potential constant at -300 mV a voltage that is realistic for both anodic and

cathodic electrodes in MFC. The other (biofilm free) electrode served merely as counter

electrode.

Figure 6.4 shows the changes of current, pH, Eh, DO and cathodic electron flow of the

potentiostatic MFC over an extended period during which the biofilm received alternating

supply of acetate or oxygen. This alternating supply of electron donor (acetate) and electron

acceptor (oxygen) inverted the electrode repeatedly from anode to cathode and back to anode. In

contrast to the operation without polarity inversion (Figure 6.3B), acetate gave repeated current

signals on a sustainable basis (Figure 6.4A).

Since acetate oxidation is a proton-liberating reaction and the electrolyte had only a low

pH buffering capacity (see composition of the medium in Section 2.1), the liquid medium (here

the working electrolyte) became significantly acidified during each anoxic (anodic) phase. For

each acetate addition (0.5 mmole), a drop of about 0.3 units of pH was observed (Figure 6.4B).

This pH drift was counteracted by increased DO level which also initiated an increase in

electrolyte Eh from around -200 to +200 mV (Figure 6.4C and D). As the potentiostat


- 109 -

maintained the biofilm working electrode at a constant potential of -300 mV the oxygen reaction

with the electrode caused now a reverse electron flow (cathodic current), which - as expected -

consumes protons at this working electrode (Figure 6.4A).

Figure 6.5 Effect of intermittent supply of acetate and oxygen on (A) current and pH; (B)

inflow/outflow dissolved oxygen concentrations and (C) cathodic electron flow rate

calculated from OUR and current in the biofilm-contained working chamber of a

potentiostatic microbial fuel cell. Note: black arrows indicate acetate addition; working

electrode potential was poised at +200 mV throughout the experiment.

By intermittently reversing polarity, the addition of acetate spikes resulted in current

peaks for longer and giving an average anodic coulombic recovery (83.5 4.6 %) for each

5.0

5.5

6.0

6.5

7.0

-10

40

90

140

190

240

pH

Cu

rre

nt (

mA

)

Current

pH

0

0.1

0.2

0.3

DO

(m

mo

l O2/

L)

Inflow

Outflow

0

2

4

6

0 5 10 15

Ca

tho

dic

Ele

ctr

on

Flo

w

Ca

lcu

late

d

(mm

ole

e- / h

)

Time (h)

From OUR


5.0

5.5

6.0

6.5

7.0

-10

40

90

140

190

240

pH

Cu

rre

nt (

mA

)

Current

pH

0

0.1

0.2

0.3

DO

(m

mo

l O2/

L)

Inflow

Outflow

0

2

4

6

0 5 10 15

Ca

tho

dic

Ele

ctr

on

Flo

w

Ca

lcu

late

d

(mm

ole

e- / h

)

Time (h)

From OUR


5.0

5.5

6.0

6.5

7.0

-10

40

90

140

190

240

pH

Cu

rre

nt (

mA

)

Current

pH

0

0.1

0.2

0.3

DO

(m

mo

l O2/

L)

Inflow

Outflow

0

2

4

6

0 5 10 15

Ca

tho

dic

Ele

ctr

on

Flo

w

Ca

lcu

late

d

(mm

ole

e- / h

)

Time (h)

From OUR


A I

R

A I

R

A I

R

A I

R

A I

R

A I

R

A.

B.

C.


- 110 -

acetate addition that was similar to that with pH control at 7 (see Figure 6.3A). These results

suggested that the anodophilic activity of the biofilm were not affected by the intermittent

presence of DO in the system.

6.3.3 Operating the Described MFC at +200 mV did not enable a Cathodic Electron

Flow

A potential of -300 mV is typical for the anodes of MFC but it is lower than what

cathodes are usually aimed to be. To test whether the above process can also operate at a higher

electrode potential, the experiment was repeated at a different potential of +200 mV (Figure 6.5).

In contrast to the experiment at -300 mV, no significant cathodic current was observed which is

in line with the general observation that cathodic oxygen reduction requires potentials lower

than +200 mV (Clauwaert et al. 2007b; Dumas et al. 2008b; Freguia et al. 2007c; He and

Angenent 2006; Rabaey et al. 2008). The presence of anodophilic bacteria could not make a

difference here.

Without an operating cathodic reaction, the intermittent aeration could not neutralize the

acidity of the working electrolyte. As a consequence, the acidity that was previously caused by

the anodic acetate oxidation of the same biofilm could not be compensated, resulting in a

gradual decrease of electrolyte pH and diminishing anodic activity of the bacteria (Figure 6.5A).

6.3.4 Cathodic Electron Balance

In analogy to the anodic coulombic efficiency, which balances the measured electron

flow against the amount of electron donor (acetate) reacted, a cathodic coulombic efficiency can

be determined from the ratio of electrons flowing away from the cathode and the amount of

electron acceptor (here oxygen) consumed. This cathodic electron balance showed that there was

always more oxygen used by the biofilm than could be accounted for by the current driven by

the potentiostat. Even with no current some background oxygen uptake activity was observed

(Figure 6.4E). This oxygen uptake is most likely due to cell maintenance respiration or the

oxidation of cell internal energy storage produced from acetate during phase 1. This is supported

by the observation that at cathodic potentials of +200 mV no cathodic current was generated at

the working electrode but significant oxygen was consumed (Figure 6.5C).


- 111 -

6.3.5 Catalytic Effect of the Anodophilic Biofilm on the Cathodic Reaction

The conclusions from the above experiments are that the cathodic reaction of the

working electrode coated with biofilm worked satisfactorily at -300 mV but not at +200 mV. In

order to establish in which range of cathodic potentials the described biofilm shows the greatest

catalytic effect, the cathodic current of the biofilm coated MFC electrode was compared to that

of a fresh, abiotic graphite electrode (Figure 6.6).

Figure 6.6 Effect of the anodophilic biofilm on the cathodic current obtained at different

electrode potentials: (A) current vs. electrode potential; (B) current enhancement factor (i

Bio/ i Abio) vs. electrode potential.

While the abiotic electrode showed the typical effect of oxygen overpotential with

significant current only being generated below -300 mV, the anodophilic biofilm enabled current

0

2

4

6

8

10

12

-450-400-350-300-250-200-150-100-50


iBio/ i

Abio

-90

-80

-70

-60

-50

-40

-30

-20

-10

0

Cu

rre

nt (m

A)

Biofilm-Electrode

Abiotic-Electrode

A.

B.


- 112 -

to be produced at -200 mV, signifying a lowering of the cathodic overpotential (Figure 6.6A).

This observation has been described before for cathodophilic biofilms (Clauwaert et al. 2007b;

Freguia et al. 2007c; Rabaey et al. 2008). At cathodic potentials between -250 and -300 mV the

presence of anodophilic bacteria enabled up to 10 times higher current than the abiotic control

(Figure 6.6B). At potentials more negative than -300 mV, the effect of the anodophilic biofilm

on the cathodic electron flow became less significant owing to the increased abiotic cathodic

current (Figure 6.6B). However, the less negative potentials (>-300 mV) are more relevant for

the practical operation of MFC.

Figure 6.7 Effect of the anodophilic biofilm on the hypothetical MFC power production: (A)

current-electrode potential curves; (B) power curves. Notes: power and voltage values

were calculated from P = V* I, where V is the difference between anodic and cathodic

0

10

20

30

40

50

60

70

80

-450 -400 -350 -300 -250 -200 -150 -100

Cu

rre

nt (

mA

)


Biofilm cathode

Abiotic cathode

Biofilm anode (refer to Cheng et al., 2008)

A.

0

5

10

15

20

25

30

0 10 20 30 40 50

Po

we

r D

en

sity (W

/m3)

Current (mA)

Biofilm cathode

Abiotic cathode

B.

Rectangle Area= Power


- 113 -

potentials; The dotted arrows extending from the dotted circles indicate the theoretical

maximal currents obtainable with the abiotic or anodophilie-catalyzed cathode.

6.3.6 Expected Power Production of MFC with Anodophilic Bacteria at both the Anode

and Cathode

The results above described the cathodic current produced by an anodophilic biofilm

coated electrode as a function of the cathodic potential. Combining this cathodic polarization

curve with the anodic polarization curve of the same cell as published earlier (Cheng et al. 2008)

allows to predict the maximum MFC power production (Figure 6.7A, B).

In an electricity producing MFC, the anodic current must be equal to the cathodic current.

Hence, the intersection of the anodic and cathodic polarization curves in Figure 7A displays the

maximum possible current (assuming zero resistance) that can be produced by a MFC with both

electrodes covered by the biofilm. For any given anodic and cathodic potential, the current and

power produced can be predicted from the rectangle areas. (V × I) as illustrated in Figure 7A.

This information also allows us to obtain the traditional power curve (Figure 6.7B). The

differences in abiotic and biofilm catalyzed cathodic polarization curves allow to estimate that,

in our case, about five times more power can be produced due to cathodic catalysis of the

biofilm.

6.3.7 Implication of the Findings

6.3.7.1 Potential Benefits of the Proposed Concept to developing Sustainable MFC

Processes

In the literature, Freguia et al. (2007c) had proposed a new operating regime to address

the limitations of pH membrane gradient and poor oxygen reduction. This was accomplished by

allowing an acidified anolyte effluent to serve as a continuous feed for the cathode. As such, the

excess protons from the anode could react directly in the cathodic oxygen reduction reaction.

Further, the residual amount of organic carbon (e.g. acetate) in the anolyte could also support the

growth of heterotrophic bacteria at the cathode. These cathode associated bacteria could further

oxidize the remaining organics in the wastewater and were capable of catalyzing the cathodic

oxygen reduction, resulting in an overall power increase without the need of an active pH

control.


- 114 -

However, as mentioned by these authors the oxygen consumption by bacteria at the

cathode in the presence of soluble organic carbons could compete for oxygen with the cathode

and diminish electricity generation (Freguia et al. 2007c). Further, the inevitable anodic biofilm

acidification would still be an issue in such a system (Rozendal et al. 2008a; Torres et al. 2008b).

In contrast to the above, our proposed MFC operating regime may offer the following

additional benefits:

The separation of acetate and oxygen addition phases may avoid the undesirable growth

of aerobic heterotrophic bacteria in the system, which may potentially out-compete the

cathodophilic bacteria in the biofilm community.

As protons are continuously liberated within an anodophilic biofilm upon substrate

oxidation, the current production can be limited by the rate at which the protons are

transported out of the biofilm (Torres et al. 2008b). Our concept of alternating flow of

electrons not only can alleviate such localized pH effect inside the biofilm but use the

protons accumulated from the anodic reaction to stimulate the subsequent proton

requiring cathodic reaction. This is because the protons are not required to be transported

out of the biofilm, but indeed the protons can reside within the biofilm for the subsequent

oxygen reduction reaction.

6.3.7.2 Anodophiles Catalyzing the Cathodic Reaction

Our observation that the same biofilm may play a role in the catalysis of both the anodic

and cathodic reaction is not entirely new. For example, Geobacter sulfurreducens, a widely

described anodophile (Bond and Lovley 2003; Dumas et al. 2008a; Fricke et al. 2008; Reguera

et al. 2006) was found to catalyze the electron transfer from the cathode to a suitable electron

acceptor (fumarate) (Dumas et al. 2008b; Gregory et al. 2004). However, fumarate is a common

electron acceptor for anaerobic bacteria such as Geobacter, and not a sustainable electron

acceptor in MFC.

In a recent study, Rabaey et al. (2008) have found that a mixed culture biofilm enriched

from an air-biocathode comprised mainly of Proteobacteria and Bacteroidetes species. A similar

microbial composition of an anodophilic biofilm was observed by Logan and Regan (2006a).

Such similarity suggests that perhaps the same groups of bacteria may catalyze both the anodic

and cathodic reactions.


- 115 -

Overall, the results in this chapter demonstrate for the first time that a mixed culture

anodophilic biofilm can also catalyze the cathodic reduction of oxygen in a single

bioelectrochemical system. The fact that an anodophilic biofilm that had developed over many

months of operation under strictly anaerobic conditions could sustain fully oxygenated

conditions without measureable side effects is significant and has also been observed recently by

Oh and coworkers (2009). The possible uses of a biofilm that catalyzes both the anodic and

cathodic reaction could lie in the operation of a traditional MFC by intermittent polarity

inversion (exchanging the electron donor and oxygen supply) at regular intervals. According to

the results of this chapter it should help overcoming the detrimental drifting of electrolyte pH

and poor cathodic oxygen reduction.

- 116 -

7 A Scalable Bioelectrochemical System for Energy

Recovery from Wastewater — Rotatable Bio-

Electrochemical Contactor (RBEC)

(A provisional patent has been filed for the invention described in this chapter)

Chapter Summary

Up to now, wastewater treatment using BES has not been successful on an industrial-

scale. This is partly due to the lack of a suitable reactor configuration for large-scale wastewater

application. In this chapter, a new BES configuration, termed rotatable bioelectrochemical

contactor (RBEC) is developed to convert the chemical energy from wastewater COD (acetate)

into electricity. Similar to rotating biological contactors (RBC), the RBEC consists of a

cylindrical water-holding vessel (ca. 3 L) which houses an array of discs (electrodes) mounted

onto a central horizontal rotatable shaft. Each disc consists of a water-immersed anodic and an

air-exposed cathodic half. No ion-exchange membrane and wastewater recirculation are required

as the air-water interface separated anode from cathode. Intermittent disc rotation (180°) enabled

each halve alternately serves as an anode and a cathode. This operating mode allows electricity

production without being affected by the pH limitation.

The COD removal caused by the action of the intermittently turning discs was increased

by about 15% (from 0.79 to 0.91 kg COD·m-3

·d-1

) by allowing an electron flow between the

anodic and cathodic disc halves. This result suggests that using suitable conductive material for

the disks of traditional RBC may already significantly improve the treatment performance.

Coupling the RBEC with a potentiostat to alleviate cathodic limitation and increase electron

flow could further increase COD removal rate to 1.32 kg COD m-3

day-1

(hydraulic retention

time 5h). About 40% of the COD degraded by the RBEC was via an electron flow meaning the

COD degrading bacteria were submersed and anoxic while the oxygen was consumed by the air

exposed cathodic disc half. While the COD removal rate was comparable to that of conventional

activated sludge processes (CAS), the RBEC removed COD more energy-efficiently than CAS

(0.47 vs. 0.7-2.0 kWh kgCOD-1

). Because of its novelty the system is expected to be improved

in both COD removal and energy use if more effective cathodic catalysts for MFC cathodes

become available.

Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor

- 117 -

7.1 Introduction

Chapter 6 has shown that a MFC anodophilic biofilm could alternately catalyze anodic

acetate oxidation and cathodic oxygen reduction reaction in a two chamber bioelectrochemical

cell with granular graphite used as the electrode material, causing a regular polarity inversion of

the system. This microbial property could alleviate both the pH splitting and poor cathodic

oxygen reduction limitations. However, it is not practical to implement the concept using the

two-chamber reactor configuration as it requires intermittent addition of a concentrated electron

donor substrate and the electron acceptor (i.e. oxygen via aeration) into the bulk liquid. This also

requires an extensive uplifting or re-circulating of the wastewater. Further, separation of anode

and cathode in the system require the use of sophisticated ionic exchange membranes, which in

general constitute a large portion of the capital cost of the whole system (Clauwaert et al. 2008a;

Rozendal et al. 2008a). Hence, eliminating the use of membranes can further reduce capital cost.

To simplify process up-scaling and to facilitate commercialization potential of BES, the

aforementioned limitations require not a small modification or process optimization but a

rethink of how the bacterial electrochemical activity can be better exploited with a more

practical and scalable BES reactor design. In this Chapter, a new BES configuration termed

rotatable bioelectrochemical contactor (RBEC) is developed. It can convert COD into electricity

without using conventional pH control methods (i.e. acid/base dosing or buffer addition) and

without the need of re-circulating the wastewater against gravity.

Similar to the well-known rotating biological contactor (RBC), the RBEC consists of a

cylindrical Perspex vessel (ca. 3 L), which houses 20 electrode discs that are mounted onto a

central horizontal rotating shaft. Each disc is divided into a water-submerged anodic and a

headspace-exposed cathodic half disc, which is partially immersed in the bulk liquid to allow

ionic conductance between the two electrodes. As such, the air-water interface served as a

―virtual membrane‖ separating the two redox zones eliminating the need of an ion-exchange

membrane. The electrodes of the RBEC are rotated such that a biofilm is established thereon and

alternately serves as the anode and the cathode by regularly swapping over the electrode discs.

The intermittent catalysis of anodic and cathodic reactions by a biofilm could help neutralize

acidity/ alkalinity of the biofilm.

The performance of the RBEC for electricity generation as a MFC is evaluated. The

feasibility of using an external power supply to facilitate the cathodic oxygen reduction and the

organic removal was also assessed. Energy consumption and treatment performances are

compared with the conventional AS process.


- 118 -


7.2.1 Construction of the RBEC Reactor

The RBEC reactor consists of three major components: (1) a reactor vessel; (2) a

rotatable electrode disc assembly; and (3) a computer controllable stepper motor unit (Figure 7.1

Photo 7.1). The reactor vessel consists of a cylindrical Perspex transparent water pipe (298 mm

length; 140 mm diameter) with its two ends covered by two separate square Perspex side Photos

(160 x 160 mm). A rubber O-ring was placed between each side Photo and the centre pipe to

assure air and water tightness.

Figure 7. 1 A schematic diagram of the RBEC system: (A) operated as a microbial fuel cell for

electricity generation (batch operation); (B) coupled with a potentiostat for enhanced

cathodic oxygen reduction operated at a chemostat mode. (1) computer monitoring and

control using LabVIEW™; (2) the RBEC reactor unit; (3) reference electrode; (4)

variable external resistor; (5) central rotating shaft; (6) electrode disc assembly; (7)

computer controllable stepper motor; (8) a computer controllable relay switch interfacing

the half-discs and the potentiostat; (9) peristaltic pump; (10) heat exchanger to warm up

the cold feed; (11) influent feed reservoir, at ~4°C ice bath; (12) effluent holding tank, at

~4°C ice bath. Note: figure not drawn to scale; RE: reference electrode; CE: counter

electrode (i.e. cathode); WE: working electrode (i.e. anode).

A rotatable electrode disc assembly consists of a Perspex horizontal central rotatable

shaft. Two separate stainless steel current collecting strips were independently mounted onto the

Potentiostat

RE CE WE

Process

Monitoring and Control

(1)(8)

(9) (9)

(11)

(12)(7)

(2) (3)

(10)

(6)

Open Air

Headspace

(5)

(A) (B)

e - e -

V

Process

Monitoring

and Control

Open Air Headspace

Ω

(1)

(2)

(6) (7)

(4)(3)

(5)


- 119 -

upper and lower sides of the shaft. 20 sets of electric conductive discs were mounted onto a

central horizontal rotatable shaft.

Photo 7.1 Photographs of the bench-scale RBEC reactor. (A) A stepper motor is used to rotate

the electrode-disc assembles in the two separate reactors; (B) a side-view of a single reactor; (C)

water level reaches the bottom edge of the upper cathodic half-disc to complete ionic circuit.

A computer controllable stepper motor unit consists of a stepper motor, a stepper motor

driver (K179 Stepper Driver), and an analog output card (Labjack U12). The motor was tightly

mounted onto the side Photo of the reactor vessel. A gear disc (one-to-seven reduction) was used

to convey the mechanical action of the motor to the central rotatable shaft.

7.2.2 Design of Two Half Discs Serving as Anode and Cathode

Each disc consists of two halves; an upper air-exposed and a lower water submerged half

(Figure 7.2, Photo 7.2). Each half consists of a stainless steel current collecting mesh.

Electrically conductive carbon fiber sheet was mounted onto both sides of the mesh. The upper

and the lower halves were physically and hence electrically separated from each other. Each half

has a small extension of the stainless steel mesh which serves as the electric contact with a

stainless steel current collector located at the central shaft. Electrical connection between the

stainless steel current collecting strips and the external circuit was completed with two

individual copper electrical wires. A Perspex plastic o-ring was used to connect each pair of

electrode discs with the central shaft and to maintain a constant distance with the neighboring

discs. The estimated total projected surface area (water submerged) of the 20 disc halves was

947 cm2. This value is used for calculating the current and power densities. After filling up the

1

2

A Stepper motor

C B


- 120 -

RBEC reactor with the wastewater, about 5 mm of the bottom edge of the air-exposed half discs

were immersed in the water. This is essential to allow an ionic contact between the air- cathode

and the submerged anode such that the system does not require the use of a membrane or

separator.

Figure 7.2 A schematic diagram of an electrode disc set. (1) A Perspex plastic o-ring; (2) a

small extension of the stainless steel mesh. Note: figure not drawn to scale.

Photo 7.2 Photographs of the rotatable electrode-disc assemble. (A) Two electrode half-discs are

mounted on a perplex O-ring; (B) stainless steel mesh enables electrical contact between

the carbon-sheet electrode and the central shaft where a stainless strip is mounted thereon;

40 mm

40 mm

120 mm

8 mm5 mm

Water level

Water Submerged

Air-Exposed

Carbon sheet

mounted on stainless steel mesh

(1)

(2)

1

2

3

Stainless-steel current collector

Copper wires external circuit

C A B

D

Rotating Shaft

E


- 121 -

(C) 20 electrode-disc assembles are mounted onto a central shaft; (D) two separate

external copper electrical wires are used to convey electrons to or from the upper and

lower discs respectively; (E) stainless steel mesh serves as a support and current collector

of the carbon sheet.

7.2.3 Process Monitoring and Control

In general, a computer program LabVIEW™ (version 7.1 National Instruments™) was

developed to continuously control and monitor the RBEC process (see Photo 7.3). Voltage

signals were recorded at fixed time intervals via the LabVIEW program interfaced with a high

precision voltage data acquisition board (DAQ) (National Instruments™ NI4350). All electrode

potential (mV) described in this study refers to values against Ag/AgCl reference electrode (ca.

+197 mV vs. standard hydrogen electrode).

Photo 7.3 Computer programs were written to control and monitor the RBEC bioprocess: (A)

RBEC operated as a MFC mode for net electricity generation; (B) RBEC coupled with a

potentiostat.

When the RBEC was operated as a MFC, the potential differences between the air-

exposed cathode and water submerged anode (i.e. cell voltage); and between the submerged

discs and a silver-silver chloride reference electrode (Bioanalytical System, RE5) were

continuously monitored by the DAQ. When a potentiostat was used to supply extra electrical

power to the system and to maintain a constant potential of the water submerged anode, the

current was directly recorded from the potentiostat.

A B


- 122 -

Rotation of the electrode discs was fully automated. The stepper motor and its driver is

calibrated such that upon receiving a proper signal from the computer program (via a Labjack™

analog output card), the electrode discs could be rotated at a defined time interval either for a

full-(360°) or a half-(180°) rotation (intermittent-flipping). In the latter case, each half disc could

intermittently serve as a submerged anode and an air-exposed cathode (Figure 7.3). When the

RBEC was coupled to a potentiostat, a computer controllable relay-switch was fabricated to

intermittently switch over the connection between the working electrode of the potentiostat and

the two half discs (see Photo 7.4). The air-exposed halves always serve as the counter electrode.

Hence, both half discs would alternately expose to an identical anodic and cathodic environment

during the intermittent-flipping operation.

Figure 7.3 Sequential disc rotation and control of the submerged anode potential during an

alternating flipping operation of the RBEC reactor.

Photo 7.4 A computer controllable relay-switch for intermittently switching over the connection

between the working electrode of the potentiostat and the two half discs.

Air

Water2..

1.Air

Water

Cathode

Air

Water1.

2.Air

Water

1st

2nd

3rd

4th

180 Clockwise

Rotation

180 Anti- Clockwise Rotation

Anode

Anode

Cathode

1.

2.

2.

1.

Connect to

RBEC Connect to

Potentiostat


- 123 -

7.2.4 Bacterial Inoculum and Synthetic Wastewater

A return activated sludge collected from a local municipal wastewater treatment plant

was used as the inoculum. It was stored at 4°C prior use. A synthetic wastewater was used

throughout the study. Unless otherwise stated, it consisted of (mg L-1

): NH4Cl 125, NaHCO3 125,

MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL L-1

of trace element

solution, which contained (g L-1

): ethylene-diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O

0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22,

NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. Sodium acetate

(5 to 10 mM) was the sole electron donor in this study. Yeast extract (0.1-1 g L-1

final

concentration) was added as bacterial growth supplement. Unless otherwise stated, the initial pH

of the synthetic wastewater was adjusted to 6.9-7.2 using either 1M HCl or 4M NaOH. No

additional pH buffer was added.

7.2.5 Reactor Operation

7.2.5.1 Start-up as a Microbial Fuel Cell

The RBEC was loaded with an activated sludge amended (10%, v/v) synthetic

wastewater (1.75 L). It was operated in batch mode as a MFC for about two months (Figure

7.1A). A variable external resistor (2 - 1M Ω) was used to connect the upper and the lower disc

half, completing the electric circuit of the system. Unless otherwise stated, the RBEC was

operated at ambient temperature (22 -25oC) and atmospheric pressure. Polarization curve

analysis was regularly conducted to evaluate MFC performance over time (Logan et al. 2006).

Briefly, it was done by decreasing the external resistance from 1M to 5 Ω in a stepwise manner.

For each external resistance setting, at least 5 min waiting period was given to obtain steady

state values of current (I) and voltage (V). These values were used to obtain the corresponding

power output (P) according to an equation of P = V × I. Power and current densities were

obtained by normalizing the P and I with the projected surface area of the submerged half discs

(i.e. 947 cm2), respectively. Correlation between the submerged anode potential and the current

density was further obtained from the data of the polarization curve analysis. This correlation is

useful to evaluate anodophilic property of the electrochemically active biofilm in a MFC (Cheng

et al. 2008). Acetate concentration of the wastewater was regularly quantified to obtain COD

removal rate at different runs (Cheng et al. 2008). Chemical oxygen demand (COD) was derived

from acetate concentration assuming 1.067 g COD·g acetate-1

.


- 124 -

After start up, COD removal rates at external resistances of 2 Ω and open circuit were

compared to test if allowing current generation in the RBEC could accelerate COD removal. The

effect of disc rotation (full-turn, half-turn or no rotation) on COD removal was also assessed at

open circuit.

7.2.5.2 Coupling the RBEC with a Power Source

After a two months operation as a MFC, the RBEC was coupled with an electrical power

source (potentiostat) to overcome any cathodic limitation in the system. The working (WE),

counter (CE) and reference electrodes (RE) of the potentiostat (Model no.362, EG&G, Princeton

Applied Research, Instruments Pty. Ltd.) were connected to the submerged anode, air-exposed

cathode and the reference electrode, respectively (Figure 7.1B; Photo 7.5). Unless otherwise

stated, the anode potential was controlled at -300 mV.

The potentiostat-assisted RBEC was initially operated in fed-batch mode for about a

week before switching into continuous mode. A synthetic wastewater (10 mM acetate) was

continuously fed through the reactor at a flow rate between 2.5 and 6 ml min-1

, corresponding to

a hydraulic retention time (HRT) between 4.9 and 11.7 h; and an organic loading rate (OLR)

between 1.3 and 3.1 kg COD·m-3

·d-1

. COD removal rate was obtained according to: (CODIn

-

CODOut

)/HRT, where CODIn

and CODOut

are COD concentrations (mg·L-1

) of the influent and

effluent, respectively. Both the influent reservoir and the effluent were stored at 4oC to avoid

microbial deterioration.

7.2.6 Scanning Electron Microscopy of Biofilm-Electrode

The morphology of the established electrode biofilm was examined by using scanning

electron microscopy (SEM) (Philips XL 20 Scanning Electron Microscope). Small pieces of

carbon electrode were removed from the reactor with a sterile stainless steel scissor. The

electrode samples were fixed in 3 % glutaraldehyde in 0.025 M phosphate buffer (pH 7.0),

overnight at 4 oC. The samples were then rinsed three times (5 min each time) with 0.025 M

phosphate buffer (pH 7.0) before they were subjected to dehydration procedure through a series

of ethanol solution (each solution was changed two times and each change was 15 min): 30, 50,

70, 90 and 100%. The 100% ethanol solution was replaced with amyl acetate (two changes, 15


- 125 -

min each. This step was performed in fume hood). The samples were critical point dried before

they were mounted to SEM specimen holder with contact adhesive. The samples were then

sputter coated with gold before being subjected to SEM observation. A fresh carbon electrode

was used as the abiotic control.

Photo 7.5 Chemostat operation of the RBEC reactor coupled with an external power supply

(potentiostat). (1) RBEC reactor; (2) potentiostat; (3) influent feed; (4) effluent tank; (5)

ice bath; (6) peristaltic pump; (7) computer for process control and monitoring.

(1)

(2)

(3) (4)

(5)

(6)

(7)


- 126 -


7.3.1 Operation as a Microbial Fuel Cell for Electricity Generation

The described reactor was constructed, inoculated with activated sludge and operated as

a fed-batch MFC reactor for about two months.

Unlike other membrane equipped, two-chamber MFC where an anode is usually

maintained under strict anaerobic condition, the absence of a membrane in the RBEC would

inevitably allow oxygen to diffuse from the air-headspace to the anodic region in the water

phase. Yet, a reasonably reducing redox state of this anodic region is essential for electricity

generation. In essence, the anode must first be polarized (bio-electrochemically) at a potential

low enough to drive electron flow towards the cathode. This could be achieved by the addition

of acetate (ca. 10 mM), amended activated sludge (10%, v/v) to the RBEC, which reduced the

anodic potential to a steady level of -453 mV at 1M Ω within 16 hours (Figure 7.4).

Figure 7.4 Effect of activated sludge on the evolution of anode potential and cell voltage of the

RBEC reactor during the initial startup. External resistance was 1M ohm. Activated

sludge was inoculated at time zero.

7.3.1.1 Increased Anodophilic Activity increases Power Output over Time

The RBEC was operated as explained under Section 7.2 by rotating full turns every 15

min. This allowed the activated sludge bacteria to colonize both, the submersed anode half and

-600

-400

-200

0

200

400

600

0 5 15 20Time (h)

Vo

lta

ge

(m

V)/

Po

ten

tia

l (m

V v

s. A

g/A

gC

l) Voltage-10% AS

Voltage-0% AS

AP-10% ASAP-0% AS


- 127 -

the air exposed cathode half of each disc. Electricity generation of the RBEC was evaluated over

50 days by using polarization curve analysis. Similar to other MFCs, the current and power

outputs of the RBEC reactor increased over time (Figure 7.5) reaching a maximum current

density and power density of 0.075 A·m-2

and 8.4 mW·m-2

respectively.

Figure 7.5 Polarization and power density curves recorded at Day 1, 14, 38 and 50. Reactor was

operated in Batch mode (acetate concentrate ranged from 5 to 11 mM).

The observed increase of current output could in theory be due to microbial enhancement

of either the cathodic or anodic half-reaction. To determine which of the two half-reactions was

stimulated by the developing biofilm, a series of current/potential curves was plotted (Figure

7.6). A clear increase in the anodic reaction could be observed from the fact that the bacteria

could generate more current for the same anodic potential. For example, at an anodic potential of

-500 mV the bacteria could generate twice as much current on day 50 compared to day 38

0

100

200

300

400

500

600

Vo

lta

ge

(m

V) Day 1

Day 14Day 38Day 50

0

1

2

3

4

5

6

7

8

9

0 0.02 0.04 0.06 0.08

Current Density (A m )-2

Po

we

r D

en

sity (m

W m

-2)

Day 1Day 14Day 38Day 50


- 128 -

(Figure 7.6). This indicates that over time the biofilm has become more capable of using the

water-immersed anode as their electron acceptor.

Figure 7.6 Current density as a function of anodic potential of the RBEC reactor obtained at

different times. The reactor was operated at batch mode with the medium saturated with

acetate.

7.3.2 In Situ Supply of Oxidizing Power via Conductive Contactor (as Electron Flow)

increases COD Removal

Although the design of the RBEC is similar to a RBC, their underlying principles of

COD removal are different. In RBC, the contactor discs are continuously rotated (1-10

revolutions per minute) to provide oxidizing power (dissolved oxygen) for the biofilm to

aerobically degrade the COD (Cortez et al. 2008; Di Palma and Verdone 2009). COD removal

can be accelerated by rotating the contactor faster to increase mass transfer of dissolved oxygen

to the biofilm (Cortez et al. 2008; Israni et al. 2002). A concentration gradient of both the COD

(electron donor) and oxygen (electron acceptor) exists along the biofilm depth, of which the

outer layer (near the bulk) always has a higher concentration compare to the inner layer (near the

contactor). While in RBEC, oxidizing power is provided in-situ from the contactor surface. The

biofilm can oxidize the wastewater COD by using the contactor directly as their electron

acceptor, alleviating the demand of disc rotation for oxygen mass transfer.

0

0.02

0.04

0.06

0.08

-600 -500 -400 -300 -200 -100


Cu

rre

nt D

en

sity (A

m -2) Day 1

Day 14Day 38Day 50


- 129 -

Figure 7.7 Changes in (A) current, (B) anodic potential and (C) acetate concentration in the

RBEC reactor operated at closed or open circuit.

To test whether by allowing electron flow from the water-immersed to the air-exposed

half disc could accelerate COD removal, COD removal rates in the RBEC operated at closed (2

Ω) or open circuit were compared. The discs were regularly rotated (1 turn at 1rpm every 15 min)

to approach RBC operation (Figure 7.7). At open circuit, the regular disc rotation could lead to a

52% increase in COD removal rate (from 0.32 to 0.48 kg COD·m-3

·d-1

) (Figure 7.7C). Similar to

the conventional RBC, the rotating action of the discs would have increased the oxygen mass

transfer and resulted in a faster microbial COD degradation. Current generation could further

increase COD removal rate by 35% (from 0.48 to 0.65 kg COD·m-3

·d-1

) (Figure 7.7C). Hence,

allowing electron flow from the water-immersed to the air-exposed half disc could accelerate

COD removal. This implicates that the treatment performance of traditional RBC may be

significantly improved by simply using conductive contactor materials.

0

1

2

3

4

5

6

Cu

rre

nt(

mA

) .

-550

-450

-350

-250

-150

-50

An

od

ic P

ote

ntial.

(mV

vs. A

g/A

gC

l)

. 2 Ohm

Open Circuit

0.43 mM·h-1

(R2 = 0.99)0.21 mM·h-1

0

1

2

3

4

5

0 5 10 20 25 30Time (h)

Ace

tate

Co

ncentr

atio

n

(mM

) .

2 Ohm (Rotated 360°Every 15 min)Open Circuit (Rotated 360°Every 15 min)Open Circuit (No Rotation)

0.32 mM·h-1

(R 2 = 0.99)

(R 2 = 0.99)


- 130 -

7.3.2.1 Parasitic Current decreases Coulombic Recovery

Although allowing current generation would favor COD removal in the RBEC, only low

Coulombic recovery was obtained (4.2%). This would most likely be due to a so-called

―parasitic current‖ effect as recently described by Harnisch and Schroder (2009). Since the

cathode discs were partially (~8% of the projected surface area of the half disc) immersed in the

wastewater to maintain ionic balance in the absence of an ion-exchange membrane, the bacteria

at the submerged region of the cathode might anodically oxidize the COD in the bulk via an

internal short circuit reaction at the cathode. As about 92% of the projected surface area of the

cathode was exposed to the air, the electrons generated at the submerged cathodic region would

readily flow towards the air-exposed counterpart and finally react with oxygen (Figure 7.8). This

short circuit current would by-pass the external circuit of the system and became unaccountable

for the Coulombic estimation. Further, the anodic reaction at the cathode would ―drag down‖ the

apparent cathodic potential to a more negative level (mixed potential). As a consequence, the

open cell voltage is diminished.

Figure 7.8 A schematic illustration of the principle of the parasitic current that could occur at

the air-exposed cathodic half disc of the RBEC. Note: diagram is not drawn to scale.

Modified after Harnisch and Schroder (2009).

Ext. Ω

Biofilm

Water

COD

CO2 + H+

O2 + 4H+

H2O

e-

e-

e-

e-

COD

O2

Parasitic Current

Anode

CathodeE°’

O2/H20

CH2O/CO2

EmixedAir


- 131 -

To evaluate the effect of parasitic current, the RBEC was operated at two different water

levels: 1) Low level where the entire cathode discs were exposed to the air. Ionic flux was

maintained by placing a non-conductive polymer sponge (as a salt-bridge) between the two

electrode halves; 2) High level where the cathode discs were partially immersed in the water as

in the original setting. The cell voltage and electrode potentials were recorded over time (Figure

7.9).

At low water level, the cell voltage was about 400 mV. Increased the water level resulted

in a gradual decrease in cell voltage to about 270 mV. Such decline was not due to the anodic

but to the cathodic half reaction, which had gradually decreased the cathodic potential from -180

to a more negative, mixed potential of -330 mV (Figure 7.9). This signifies the presence of a

parasitic current at the cathode, which represents an alternative path of COD removal even at

open circuit operation (Figure 7.10). From an electricity generation standpoint, the parasitic

current is undesirable as it diminishes the Coulombic recovery. However, it would still be

desirable from a wastewater treatment standpoint that the bacteria can degrade COD using the

conductive contactors of the RBEC at either close or open circuit.

Figure 7.9 Effect of water level on the electron potentials and cell voltage of the RBEC under

open circuit. Notes: A non-conductive sponge was sandwiched between the anodic and

cathodic discs to allow ionic flux during low water level operation; reactor operated at

batch-mode with acetate saturation (>2mM). pH 8.5 at 35oC; the discs were rotated for a

full turn once per hour at 1 rpm.

-600

-400

-200

0

200

400

0 5 10 15 20 25 30Time (h)

Cell

Voltage (

mV

)/

Ele

ctr

ode P

ote

ntial (m

V v

s. A

g/A

gC

l)

Cell Voltage

Cathode

Anode

Open Circuit

Cathode

Anode

Low Cathode

Anode

High


- 132 -

Figure 7.10 Effect of external current on (A) acetate removal and (B) electrode potentials.

RBEC operated at batch mode with the cathodic half discs partially immersed in the

water, pH 8.5 at 35oC; the discs were rotated for a full turn once per hour at 1 rpm.

7.3.3 Sequential Flipping the Electrode Discs allows Alternate Current Generation

Chapter 6 suggests that a BES biofilm could catalyze both anodic substrate oxidation and

cathodic oxygen reduction in a two-chamber BES, causing an intermittent polarity inversion of

the system. Such bidirectional microbial electron transfer could alleviate both the pH gradient

and cathodic oxygen reduction limitations in the BES (Chapter 6). The same concept was tested

here with the RBEC. The discs were intermittently flipped (180°, every 1h) instead of fully

rotated such that each half disc was alternately exposed to the wastewater and the air (see Figure

7.3). The result showed that the intermittent-flipping action could revert the polarity of current

(ranged from -6.7 to +7.5 mA), corroborating with the finding that the biofilm could alternately

catalyze the current under alternating redox conditions (Figure 7.11A). An averaged power

density of 0.93 mW·m-2

was obtained. In terms of COD removal, allowing current generation

under the intermittent-flipping operation could accelerate the COD removal rate by 15%

compared to the open-circuit control (from 0.79 to 0.91 kg COD·m-3

·d-1

) (Figure 7.11B). This

-600

-500

-400

-300

0 5 10 15 20 25 30Time (h)

Po

ten

tia

l (m

V v

s. A

g/A

gC

l)

0

5

10

15

20

Cu

rre

nt (

mA

)

0

1

2

3

4

5

Ace

tate

(mM

)

Open circuit Close circuit (2Ω)

0.20 mM∙h-1

0.11 mM∙h-1

Cathode

Anode

Acetate

(B)

(A)


- 133 -

again suggests that allowing electron flow from the water-immersed to the air-exposed half disc

could accelerate COD removal.

Figure 7.11 Effect of regular half-rotation of electrode dices on (A) current, anodic potential and

(B) acetate concentration in the RBEC reactor operated at close (2 ohm) and open circuit.

The discs were flipped (180°) every 1 hour. Solid arrows in A indicate flipping events

(~1min).

Interestingly, when current was prohibited (open circuit) the intermittent-flipping

resulted in a 2.5-fold increase in COD removal rate compared to the control without rotation

(from 0.32 to 0.79 kg COD·m-3

·d-1

). Since the biofilm was alternately exposed to the anoxic

wastewater and oxygen, under oxygen limited condition (i.e. submerged in the wastewater) the

biofilm may undertake storage of substrate (acetate) from the wastewater as intracellular

polymers such as polyhydroxybutyrate (PHB) (Van Aalst-van Leeuwen et al. 1997; Van

Loosdrecht et al. 1997). Subsequent exposure to the air may allow the bacteria to oxidize the

storage materials with oxygen. Hence, apart from the parasitic current effect, bacterial storage

-10

-5

0

5

10

Cu

rre

nt (m

A)

0.60 mM·h-1

0.52 mM·h-1

(R2

= 0.99)

0

2

4

6

8

10

0 5 10 20 25 30 35Time (h)

Ace

tate

Co

nce

ntr

atio

n

(mM

)

Open Circuit

2 Ohm

Half-Turn (180°) Disc Rotation Every 1h:

(R2

= 0.95)

(A)

(B)


- 134 -

may account for the enhanced COD removal under the intermittent-flipping operation of the

RBEC. Yet, further study is required to clarify this point.

7.3.4 Coupling the RBEC with an External Power Source to achieve Higher Current

Similar to most other air-cathode based MFC, the sluggish cathodic oxygen reduction has

also limited the electricity generation in the RBEC (Clauwaert et al. 2009; Kim et al. 2007;

Wang and Han 2008). As a result, the anodic half discs became an unfavorable electron acceptor

for the biofilm (i.e. very negative potential) even when the external resistance was not limiting.

In the literature, cathodic limitation in a BES could be overcome by providing an extra voltage

to ―boost‖ the cathodic reaction. A well known example is cathodic hydrogen production using

MEC (Logan et al. 2008). Here, a potentiostat was used to overcome the cathodic overpotential

and to keep a constant, more suitable (less negative) anode potential (-300 mV) for the biofilm

to ―respire‖. However, instead of generating hydrogen under anaerobic condition the cathode

remained as oxygen based.

7.3.4.1 Electrochemically Assisted Anode Facilitates the Establishment of Anodophilic

Biofilm

The RBEC was coupled to a potentiostat as described in the method section. In the

absence of acetate, a background current of ~ 2 mA and a cathodic potential of -300 mV were

obtained (Figure 7.12A, B). The acetate addition resulted in a similar current profile as observed

in the previous, non-potentiostatic assisted mode but at about 10-fold higher current (about 60

mA) (Figure 7.12A). The increased current was caused by the potentiostat, which has provided

the additional potential to overcome the cathodic reduction limitation. In this example where the

anode was maintained at -300 mV, a cathodic potential of about-1800 mV was required to

substantiate the current triggered by the anodic substrate oxidation (Figure 7.12B). A higher

Coulombic recovery of 40% (versus ca. 5% in non-assisted mode) was obtained, indicating that

the biofilm has become more anodophilic with the more ―attractive‖ anode.

After a several fed-batch cycles, the RBEC was switched into a continuous mode to

avoid substrate limitation (Figure 7.12C, D). The current was maintained at a similar level

between 40 and 60 mA as observed during the batch-mode operation. The average COD


- 135 -

removal rate was 1.15 ± 0.05 kg COD·m-3

·d-1

. Overall, the anodic biofilm has become more

anodophilic with a more suitable anodic potential.

Figure 7.12 Performance of a potentiostat-controlled RBEC operated at (A, B) batch mode and

(C, D) continuous mode. The submerged anode discs were poised at -300 mV vs.

Ag/AgCl. The discs were rotated (360°) once per 15 min.

Figure 7.13 Current-potential plots suggest that the two half discs have (A) similar cathodic

activity, but (B) different anodophilic activity. WE: working electrode; CE: counter

electrode.

0

20

40

60

80

Cu

rre

nt

(mA

)

48 Time (h) 84 9672

Anode

Cathode

1.10 kg COD·m-3·d-1

1.19 kg COD·m-3·d-1

1.16 kg COD·m-3·d-1

Ac

eta

te S

tarv

ed

(Ba

tch

mo

de

)

Acetate Saturated

(Continuous mode)

0

20

40

60

Cu

rre

nt

(mA

)

1

2

3

4

5

6A

ceta

te (

mM

)

-1800

-1400

-1000

-600

-200

0 12 20 3624

Po

ten

tia

l (m

V v

s.

Ag

/Ag

Cl)

Anode

Cathode

-1800

-1400

-1000

-600

-200

Acetate

(A) (C)

(B) (D)

0

20

40

60

-800 -400 0 400 800

Working Electrode (WE) Potential (mV vs. Ag/AgCl)

-60

-40

-20

0

-1500-1000-5000

Cu

rre

nt (m

A)

Half discs mostly exposed to air Half discs mostly submerged in water

(A): (B):WE

CE

CE

WE


- 136 -

Prolonged operation with periodic full disc rotation resulted in a difference in the

bioelectrochemical properties between the two half discs as each half would serve

predominantly as either an anode or a cathode. The relationship between the current and

electrode potential indicates that the two half discs exhibited similar cathodic but different

anodic properties (Figure 7.13).

7.3.4.2 Sequential Flipping the Electrochemically-Assisted Discs establishes

Anodophilic Biofilm on both Half Discs

In order to establish a suitable biofilm capable of catalyzing both an anodic acetate

oxidation and cathodic oxygen reduction on both half discs, the process was modified such that

the disc was only rotated for half a turn (180 degree) instead of a full turn (360 degree). Hence,

each half disc could serve as a submerged anode and an air-exposed cathode intermittently. The

submerged anodic half discs were always controlled at fixed potential of -400 mV (Figure 7.14).

Figure 7.14 Intermittent Disc-flipping operation of the RBEC with the submerged anodic discs

controlled at a potential of -400 mV. (A) Anodic current over time; (B) electrode disc

potential over time; Prior to the experiment, the biofilm at disc 2 was more anodophilic

than that at disc 1 (Figure 7.13). In the first 16 h the discs were flipped once every 30

min, thereafter they were flipped once per hour.

0

10

20

30

40

50

Cu

rre

nt (

mA

)

-1200

-1000

-800

-600

-400

0 24 48 72 96Time (h)

Po

ten

tia

l (m

V v

s. A

g/A

gC

l)

Fixed Anodic Potential

1

2

2

1

Intermittent Flipping cathode

anode

cathode

anode

Cathodic Potential

(A)

(B)


- 137 -

Initially, anodic current was not observed when disc 1 (the half discs mostly exposed to

the air in previous runs) served as the anode because of the lack of anodophilic activity as

suggested in Figure 7.13. Flipping over the discs to revert the polarity could enable anodic

current of about 30 mA. This current was catalyzed by the anodophilic biofilm at disc 2 (the half

discs mostly submerged in water in previous runs) (Figure 7.14A).

Operating the RBEC for a longer period under the intermittent-flipping regime gradually

increased the anodic current catalyzed by the biofilm at disc 1 (Figure 7.14A). This current

improvement also coincided with an increase in COD removal rate over time. At 51, 61 and 92 h

the COD removal rates were 0.60, 0.88 and 1.31 kg COD·m-3

·d-1

, respectively. Therefore, the

intermittent flipping of disc and polarity inversion could establish electrochemically active

biofilm at both half discs, contributing to a higher COD removal from the wastewater.

7.3.5 Intermittent Flipping of the Discs avoids the Continuous Alkalization of the

Cathode

The cathodic oxygen reduction consumes protons according to equation: O2 + 4e- + 4H

+

→ 2H2O. Since the protons and hydroxide ions are always in equilibrium with each other

through the water dissociation (Kw=[H+][OH

-] ≈ 10

-14), net consumption of proton represents a

hydroxide ion production: O2 + 2H2O + 4e- → 4OH

-. Hence, it is expected that the localized pH

at the cathode disc would increase over time. According to Nernst equation, a 10 fold increase of

hydroxide ion (i.e. ΔpH = +1) causes a reduction of cathodic potential by 59 mV. This

represents a loss of electromotive force (e.m.f) in the system and hence more energy is required

to sustain the forward cathodic reaction. In the electrochemically-assisted RBEC, this extra

energy is compensated by the external power source (potentiostat). Thus, neutralizing the

cathode alkalinity could reduce the electricity input in the process.

The effect of intermittent flipping of the electrode discs on the cathode pH was evaluated

(Figure 7.15). Before a flipping event (from 5 to 65 min), the cathode pH (disc 1) increased from

7.0 to 9.0 and the cathode potential became more negative (-74 mV/pH). A similar trend was

observed for another half disc (disc 2) from 65 to 125 min. During this period, disc 1 was

immersed in the wastewater (pH about 7) serving as an anode and its previously established

alkalinity was neutralized. After the next flipping (125 min), the cathode pH and potential


- 138 -

resumed to their original levels (pH 7 and -1270 mV, respectively). The continuous build-up of

alkalinity at the air-cathode could thus be avoided.

Figure 7.15 Effect of intermittent flipping of the electrode discs on the pH of the air-exposed

cathode. The discs were flipped at about 5, 65 and 125 min.

7.3.6 The Established Biofilm could catalyze a Cathodic Oxygen Reduction

After the biofilm was acclimatized in a RBEC with intermittent exposure to the air and

with alternating current generation, the catalytic property of the established biofilm electrode for

cathodic oxygen reduction was compared with a control (abiotic) electrode using voltammetric

technique (LSV) (Figure 7.16). In a potential range between -0.2 and -1.0 V, the biofilm

electrode outperformed the control electrode with respect to current generation. At -0.5 V the

current density of the biofilm electrode was -0.24 A m-2

, which was about 2 fold higher than that

of the abiotic control (-0.12 A m-2

). However, the cathodic overpotential of oxygen reduction

remained significantly large, and the presence of biofilm did not appear to be beneficial on

alleviating the overpotential of the cathodic reaction. This result is in contrary to the previous

7

8

9

Ca

tho

de

pH

0

50

100

150

200

250 Cu

rren

t (mA

)

Cathode pH Current

-2000

-1500

-1000

-500

0

0 30 90 120Time (min)

Po

ten

tia

l (m

V v

s. A

g/A

gC

l) .

Submerged Discs (Anode)

Air-Exposed Discs (Cathode)


- 139 -

chapter where an anodophilic biofilm could significantly reduce the overpotential of cathodic

oxygen reduction at a granular graphite electrode.

Figure 7.16 Effect of biofilm on the cathodic oxygen reduction. Potential scan rate 1.0 mVs-1

,

Current density was calculated based on a projected surface of the carbon electrode of 7

cm2, dissolved oxygen concentration in the medium was maintained at >6 mg/L.

The morphology of the same biofilm electrode was examined with SEM technique

(Figure 7.17). Compared to the abiotic control (A), a developed biofilm was established on the

electrode fibers. However, a boarder examination of the samples reveals that not all of the fiber

surfaces were colonized by the biofilm (Figure 7.17C).

Figure 7.17 Scanning electron micrographs of (A) a plain carbon fiber sheet; and (B to D)

carbon fiber sheet attached with the biofilm obtained from the RBEC under alternate disc

flipping operation (aerobic headspace).

1.0

1.2

1.4

1.6

1.8

2.0

-1-0.8-0.6-0.4-0.20


Cu

rre

nt

De

nsity (

A m

-2)

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

Ra

tio o

f Bio

/ Ab

iotic

Cu

rren

t

Ratio Bio/Abio

Biofilm Electrode

Abiotic Electrode

1

2

3

A B

C D


- 140 -

7.3.7 Energy Evaluation of the Electrochemically Assisted RBEC process

The treatment performance and energy requirement of the process are compared with

that of conventional AS processes (Table 1). The COD removal rate achieved by the RBEC fall

within the range of the conventional AS processes, suggesting that the RBEC process is equally

good as compared to the AS processes in removing organic pollutants from the wastewater. In

terms of energy usage, the RBEC process demonstrated better efficiency (0.47 kWh kg COD-1

)

compared to the AS processes (0.7-2.0 kWh kg COD-1

). Further, the RBEC process also

demonstrated at least 5 times less energy demand per volume of wastewater treated as compared

to the AS processes (88 vs. 430 kWh ML-1

) (Table 1). These figures indicate that a substantial

energy saving may be achieved if the wastewater COD was degraded via electric conductor as

electron flow instead of via forced-aeration of wastewater as in the conventional AS processes.

Table 7.1 Treatment performance and energy requirement of the RBEC process and

conventional activated sludge (AS) processes.

Conventional AS

Processes RBEC Process

a

#COD removal rate (kg COD m

-3 day

-1) 0.5 - 2

b 1.32

Energy input per COD removed (kWh kg COD-1

) 0.7 - 2 c

¥0.47

Energy per volume of wastewater treated (kWh ML-1

) 430 - 940 d

¥88

a Values for the RBEC process are calculated based on data obtained in Figure 7.14 (91-100 h).

b Adapted from (Logan et al. 2006)

c Adapted from (Tchobanoglous et al. 2003)

d Adapted from (Keighery 2004).

#COD content of acetate was estimated assuming 1.067 g COD per 1.0 g acetate.

¥The values include the energy consumption by the stepper motor (0.083 Wh per one rotation).

7.4 Concluding Remarks

Overall, this chapter has demonstrated the working principle of the proposed RBEC

process as a BES. Based on the findings, the following points are summarized:

The RBEC can operate as a MFC, achieving net electricity recovery while removing organic

pollutants from the wastewater. Even without adjusting the pH of the wastewater, the

detrimental problem of pH split as observed in membrane-BES was not encountered.


- 141 -

Intermittent disc flipping and polarity inversion establishes anodophilic/ cathodophilic

biofilm at both half discs, leading to increasing pollutant removal rate.

The RBEC bioconversion rate could be increased by controlling the anodic potential with an

external power source (potentiostat).

Intermittent flipping of the discs avoids the continuous alkalization of the cathode.

The biofilm could catalyze a cathodic oxygen reduction. However, the cathodic

overpotential of oxygen reduction remained significantly large.

The electrochemically assisted RBEC process can remove COD more energy efficient than

the conventional activated sludge process.

Treatment performance of traditional RBC process may be significantly improved by simply

using conductive contactor materials.

Further process optimization is required to maximize the functionality of the RBEC to

recover useful energy from wastewaters. These efforts may include: (i) optimizing current

density and COD removal by increasing the electrode surface area (e.g. number of contactor

discs) per liquid volume ratio; (ii) minimizing overpotential of the cathodic oxygen reduction by

modifying the electrode with a suitable catalyst or using more effective electrode materials; (iii)

apart from the direct production of electricity, the possibility of the RBEC for the production of

other valuable products or energy carriers such as hydrogen or methane gas should also be

explored.

- 142 -

8 A Rotatable Bioelectrochemical Contactor (RBEC)

enables Electrochemically Driven Methanogenesis

Chapter Summary

In this chapter, the RBEC prototype was operated as a microbial electrolysis cell (MEC)

under fully anoxic condition. The capacity of the system to couple anodic acetate oxidation with

a cathodic production of energy carrier gases (e.g. hydrogen or methane) was evaluated.

When the oxygen (air) in the headspace was replaced with nitrogen, hydrogen was

produced. However, prolonged anoxic operation resulted in the production of methane (≥50%,

v/v) instead of hydrogen (~0.1%). Such methane production was directly proportional to both

the current and the electrical energy input. For example, increased the current from 15 to 390

mA (0.16 and 4.1 A m-2

) resulted in a 7-fold increase in the COD removal rates (from 0.2 to

1.38 kg COD m-3

day-1

) and a 13-fold increase in the methane production rate (from 0.04 to 0.53

L L-1

day-1

). Over 80% of current was recovered as methane. Using the described RBEC to

convert COD (here acetate) into methane could allow energy saving and possible energy

recovery from the wastewater (-1.5 to 1.5 kWh kg COD-1

).

Although the proposed method of methane production is unlikely to replace traditional

anaerobic digestion, the fact that the bioconversion and methane generation rates could be

controlled electrochemically appears promising. Further, the absence of precious chemical

catalysts (e.g. platinum), ion exchange membranes and the low requirement of liquid

recirculation may represent reduction in capital cost, facilitating process up-scaling.

Overall, the RBEC described in this chapter may represent an alternative option for

large-scale BES-based technology for energy recovery from wastewater. The use of RBEC in

other application niches should also be explored in future.

Chapter 8: RBEC enables Electrochemically Driven Methanogenesis

- 143 -

8.1 Introduction

The use of an external power source (Chapter 7) to facilitate the cathodic oxygen

reduction in the RBEC has forfeited the possibility of generating net electrical energy as a by-

product. To increase the commercial promise of the proposed RBEC configuration, the process

should aim at converting the organic waste into a valuable by-product such as hydrogen.

Generating hydrogen from the cathode of the RBEC can be easily achieved with its enclosed

headspace design. Hydrogen should by theory be produced when a suitable cathodic potential is

provided by an external power source under anoxic condition. However, the operation of

membrane-less MEC for hydrogen production is always associated with the generic problems of

low hydrogen purity and the risk of hydrogen loss through hydrogenotrophic methanogenesis

(Call and Logan 2008; Clauwaert et al. 2008b; Logan et al. 2008). Different mitigation methods

have been proposed to optimize hydrogen yield by suppressing methane formation, but none of

these methods could eliminate methane formation in MEC.

In fact, methane formation in membrane-less MEC configurations is a very robust

process. Completely eliminating methane formation is therefore a very challenging task. On the

other hand, eliminating hydrogen formation to facilitate a high rate conversion of organics into

methane seems to be an easier option. From an energy recovery standpoint, methane could be a

more suitable energy carrier compare to hydrogen as it has a higher volumetric energy density.

Furthermore, compared to methane hydrogen has a much lower density and boiling point, only

0.09 kg Nm-3

and -252.9 °C vs. 0.717 kg Nm-3

and -182.5 °C for methane. These physical

properties make hydrogen gas unsuitable for transport and storage compared to methane, which

already has existing infrastructure to deal with its mass production. Hence, methanogenesis

could be an alternative option for practical operation of MEC.

In contrast to the previous chapter where the overpotential of oxygen at the cathode

severely limited the reaction, air was not supplied to the gas phase of the reactor such that the

cathodic halves of the disks of the RBEC reduce protons to hydrogen. However, it has been

shown in the literature that biofilms in the presence of hydrogen producing electrodes enrich for

the development of methanogenic micro-organisms (Ajayi et al. 2010; Clauwaert and Verstraete

2009; Jeremiasse et al. 2010). As the inhibition of methanogens is practically cumbersome and

the volumetric energy content of methane is higher than of hydrogen, this chapter aims at

producing methane as a byproduct from the treatment of dilute synthetic wastewater using the

RBEC.


- 144 -


8.2.1 Bacterial Seeding Inoculum and Synthetic wastewater

A return activated sludge collected from a local domestic wastewater treatment plant

(sequencing batch reactors activated sludge process) was used as the initial inoculum. It was

stored at 4°C prior use. A synthetic wastewater was used throughout the study. Unless otherwise

stated, it consisted of (mg L-1


FeSO4·7H2O 6.25, and 1.25 mL L-1


): ethylene-

diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99,

CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014,

and NaWO4·2H2O 0.050. Sodium acetate (5 to 10 mM) was added as the electron donor. Yeast

extract (1 g L-1

final concentration) was added as bacterial growth supplement. Unless otherwise

stated, the initial pH of the synthetic wastewater was adjusted to 6.9-7.2 using either 1M HCl or

4M NaOH.

8.2.2 Anoxic Operation of the Reactor Headspace of the Electrochemically-

Assisted RBEC

The enclosed headspace design of the RBEC reactor offers opportunity to recycle,

recover or control of gaseous byproducts. It is possible that by maintaining the headspace under

anoxic condition and applying a suitable external electrical power in a way similar to MEC

processes, the cathode of the RBEC reactor can generate hydrogen gas, an added-value

byproduct that is of commercial interest. All settings are similar to the system described in

Section 7.2.4.2 in Chapter 7, except that the headspace was completely isolated from the

atmospheric air throughout the experiment (Figure 8.1) (Photo 8.1). To obtain anoxic condition

in the headspace, the headspace was purged with pure nitrogen gas (1 L min-1

) for at least 10

min. In general, the influent has a composition as described in Section 7.2.3. It was continuously

fed into the reactor at a flow rate ranged from 1 to 3 ml min-1

. These flow rates corresponded to

hydraulic retention time (HRT) ranged from 9.7 to 19.4 h. COD removal rate was calculated

according to: (CODIn

-CODOut

)/HRT, where CODIn

and CODOut

are COD (mg L-1

) of the

influent and effluent, respectively. COD was measured according to a standard method (Closed

Reflux Colorimetric Method 1992)


- 145 -

Figure 8.1 A schematic diagram of the anoxic RBEC system for hydrogen or methane gas

production operated at a chemostat mode. (1) computer monitoring and control using

LabVIEW™; (2) gas counter ; (3) a computer controllable relay switch interfacing the

half-discs and the potentiostat; (4) peristaltic pump; (5) influent feed reservoir, at ~4°C

ice bath; (6) heat exchanger to warm up the cold feed; (7) the RBEC reactor unit; (8) gas

purging or sampling port; (9) reference electrode; (10) rotatable shaft and electrode disc

assembly; (11) computer controllable stepper motor; (12) effluent holding tank, at ~4°C

ice bath. Note: figure not drawn to scale.

Photo 8.1 Anoxic RBEC reactor for hydrogen or methane gas production operated at a

continuous mode. Heating tubing coil surrounding the reactor vessel was used to

maintain a constant temperature of 35°C.


- 146 -

8.2.3 Gas Production and Measurements

Biogas production from the RBEC was continuously recorded with a gas counter

connected to the headspace of the reactor. Prior to each experiment, the gas counter was

calibrated by injecting a known volume of N2 into the reactor to determine the gas volume per

count. This allows the calculation of the biogas production rate (BPR) (L biogas L-1

reactor day-

1). CH4, H2 and CO2 volumetric fractions (v/v, %) in the headspace were determined using a

Varian Star 3400 Gas Chromatograph (GC) coupled with a thermal conductivity detector. The

carrier gas was high purity nitrogen at a flow rate of 30 ml min-1

. Separation of different gas

species in the sample was performed on a Pora-PakQ packed column (2m x 5mm (internal

diameter)) maintained at 40oC. Detector and inlet temperatures were maintained at 40 and

120°C, respectively. Gas sample was taken from the reactor headspace via a gas-tight rubber

bung using a 10-mL plastic disposable syringe. About 50-100 µl of the gas sample inside the

plastic syringe were manually injected into the GC using a 100 µl gas-tight syringe (Hamilton).

Standard curves of each measured gas were established by injecting a known volume of high

purity standard gas into the GC.

The specific methane or hydrogen production rate (MPR or HPR) (L CH4 or H2 L-1

reactor day-1

) were calculated from the BPR and the volumetric fractions (v/v, %) of each gas in

the headspace (CH4headspace

or H2headspace

) according to the following equations:

MPR =CH4

headspace

100% × BPR (e.q. 8-2)

HPR =H2

headspace

100% × BPR (e.q. 8-3)

8.2.4 Examination of Electrode Biofilm using Scanning Electron Microscopy

The morphology of the biofilm growth at the electrode was examined by using scanning

electron microscopy (SEM) (Philips XL 20 Scanning Electron Microscope). At appropriate time

points as specified in the text, a small piece of carbon sheet was cut with a sterile stainless steel

scissor. The electrode samples were fixed in 3 % glutaraldehyde in 0.025 M phosphate buffer


- 147 -

(pH 7.0), overnight at 4 oC. The samples were then rinsed three times (5 min each time) with

0.025 M phosphate buffer (pH 7.0) before they were subjected to dehydration procedure through

a series of ethanol solution (each solution was changed two times and each change was 15 min):

30, 50, 70, 90 and 100%. The 100% ethanol solution was replaced with amyl acetate (two

changes, 15 min each. This step was performed in fume hood). The samples were critical point

dried before they were mounted to SEM specimen holder with contact adhesive. The samples

were then sputter coated with gold before being subjected to SEM observation. Abiotic plain

carbon sheet were treated with the same procedures and served as the Control.


- 148 -


The described RBEC was constructed, inoculated with activated sludge and operated

with an air-cathode for about three months. To establish a good anodic biofilm, a potentiostat

was connected to maintain the anodic potential at a suitable level (-300 to 0 mV) for the

anodophilic bacteria. Periodic half-turn rotation of the shaft enabled each electrode half disc to

intermittently serve as the submerged anode and the air-exposed cathode. This established a

biofilm capable of catalyzing the anodic as well as cathodic reaction (Chapter 7).

To test whether the developed reactor could also produce hydrogen as useful energy, the

oxygen in the headspace was replaced by nitrogen and its effect was recorded.

8.3.1 Anoxic Cathode enables Hydrogen Gas Formation

Switching the system from an oxygen reducing cathode (air in cathodic chamber) to an

anoxic cathode did not affect the electron flow (Figure 8.2). The potentiostat maintained a

constant anodic potential of -200 mV and hence the electron delivery by the bacteria to the

anode also stayed constant. However, switching from aerobic to anoxic headspace immediately

decreased the cathodic potential from about -1100 to -1300 mV (Figure 8.2B). Resuming the

headspace composition to air immediately increased the cathodic potential to the original level.

Under anoxic condition, the electrons at the cathode can only be accepted by electron

acceptors other than oxygen. In the literature, protons are commonly observed as the terminal

electron acceptor under such anoxic condition, leading to the cathodic hydrogen formation

according to equation 8-4:

2H+ + 2e

- → H2 E

o‘ = -612 mV (pH 7, 25°C) eq. 8-4

O2 + 4H+ + 4e

- → 2H2O E

o‘ = +618 mV (pH 7, 25°C) eq. 8-5

The reduction potential of this reaction is -612 mV while the reduction potential of the

oxygen reduction reaction is +618 mV (equation 8-5) (assume pH 7, 25°C). The results suggest

that the hydrogen producing reaction in the RBEC has about a 2.5 times lower overpotential

compared to the cathodic oxygen reduction.


- 149 -

Figure 8.2 Effect of different compositions in the reactor headspace on (A) current and the (B)

cathode potential (counter electrode potential). Submerged anode discs were poised at -

200 mV vs. Ag/AgCl. No flipping of electrode disc throughout the experiment. Solid

arrows indicate the onsets of purging with the test gas at 1 L min-1

. N2= 100% nitrogen;

CO2:N2= 20:80% carbon dioxide and nitrogen mix.

Cathodic hydrogen formation instead of oxygen reduction requires a cathodic potential

that is 200 mV lower. This would incur an increased power input in the process. However, with

hydrogen an additional fuel is produced, that may potentially offsets some of the electrical

energy input.

After switching the headspace from aerobic to anaerobic, hydrogen generation in the

RBEC was monitored at different applied anodic potentials, 0, -200 and -450 mV (Figure 8.3).

Over an 80 h period, each electrode half disc served alternately as an anode or a cathode upon an

hourly half rotation of the electrode disc assembly. The result shows that hydrogen production

was directly proportional to the current density (Figure 8.4). This observation is in line with

others using MECs for hydrogen production (Logan et al. 2008).

0

20

40

60

80

Cu

rre

nt (m

A)

-1500

-1300

-1100

-900

-700

-500

-300

0 10 20 30 40 50 60

Time (min)

Po

ten

tia

l (m

V v

s. A

g/A

gC

l)

.

Submerged Anode

Air-Exposed Cathode

Air Air AirCO2:N2 N2N2

CO2:N2


- 150 -

Figure 8.3 Effect of anode potential on (A) hydrogen production; current and (B) electrode

potentials in the RBEC operated under anoxic cathodic condition. The system was

switched from an air-cathode to anoxic cathode for about two days prior to the

experiment. Discs were flipped once per hour. Submerged disc = 1 and 0 indicate half

disc 1 and 2, respectively.

Figure 8.4 Hydrogen production rate as a function of current density in the RBEC operated

under anoxic cathodic condition. Data are calculated from Figure 8.3.

y = 0.04x - 0.01

R2 = 0.99

0.00

0.02

0.04

0.06

0.0 0.5 1.0 1.5 2.0


)

Hyd

rog

en

Pro

du

ctio

n R

ate

.

(m3 m

-3 d

-1)

(A)

(B)


- 151 -

Table 8.1 Effect of anodic potential set point on the reaction rate and net energy input of the

RBEC reactor.

Poised Anodic Potential

I COD Removal Rate

H2 Production Rate

Electricity Input per COD Removed

Energy Recovered as H2 per COD Removed

Net Electricity Input per COD Removed

a

mV mA kg ·m-3

·d-1

m3·m

-3·d

-1 kWh·kg

-1 kWh kg

-1 kWh·kg

-1

0 90 0.35 0.06 2.66 0.45 2.21

-200 72 0.26 0.05 2.12 0.50 1.62

-450 41 0.17 0.003 0.75 0.05 0.70

Data were obtained during the first 4 days after switching from aerobic to anaerobic cathodic headspace.

Energy content of hydrogen is based on its heat of combustion value, -285.8 kJ·mol-1

aOnly the energy input from the potentiostat is considered here.

On the one hand the provision of external electrical energy to ―push the reaction‖ would

be an operating cost factor, on the other hand it speeds up the reaction and hence would enable

to use smaller vessel sizes (lower capital costs). To visualize the trade-off between energy input

and reaction rate the reactor was operated at three different anodic potentials (Table 8.1). Results

show that a significant energy cost would be incurred when operating the reactor and the COD

removal rates are lower than that obtained from activated sludge plants, 0.5-2 kg COD·m-3

·d-1

(Logan 2008).

8.3.2 Methane instead of Hydrogen was the Predominant By-product after

Prolonged Anoxic Operation

After operating the RBEC reactor for over two weeks under fully anoxic condition, up to

48% of methane was detected at the headspace instead of hydrogen (0.1%). SEM photos of the

biofilm electrode collected from the methane-producing RBEC illustrated that a well developed

biofilm was developed at the carbon electrode compared to the abiotic control (Figure 8.5).

Methane generation was also observed by other authors and could not be avoided in membrane-

less system (Cheng et al. 2009b; Clauwaert and Verstraete 2008; Hu et al. 2008). Rather than

attempting to suppress methanogenesis it was aimed at using and maximizing this reaction and

produce methane to offset the energy input.

As observed for hydrogen the rate of methane formation of the RBEC was directly

proportional to both the current and the electrical energy input (Figure 8.6A and B). With an

average current density of 4.1 A m-2

(390 mA) (the anodic and cathodic potentials were -200 and


- 152 -

-1600 mV, respectively), the maximal COD removal rate and methane production rate were 1.38

kg COD m-3

day-1

and 0.53 L L-1

day-1

, respectively (Figure 8.6A). However, when the current

was reduced to less than 0.16 A m-2

(15 mA) (the anodic and cathodic potentials were -560 and -

850 mV, respectively), the COD removal rate and methane production rate were remarkably

reduced to 0.2 kg COD m-3

day-1

and 0.04 L L-1

day-1

, respectively (Figure 8.6A).

Figure 8.5 Scanning electron micrographs of (A to E) a carbon fiber sheet attached with the

biofilm obtained from the methane-producing RBEC under alternate disc flipping

operation (anoxic headspace) at different magnification; and (F) a plain carbon

fiber sheet. The samples were collected after about two weeks the reactor was

switched from aerobic cathode into fully anoxic operation. Dotted circles in A to D

indicate the same location of the same sample.

A B

C D

E F

5 µm

50 µm 5 µm

500 µm 200 µm

50 µm


- 153 -

Figure 8.6 Methane production rate and COD removal rate as a function of (A) current and (B)

electricity input in the RBEC operated under anoxic condition. Temperature was

controlled at 35°C.

When voltage was not applied and the external circuit was shorted (i.e. close to zero

resistance between the upper and lower half discs), the methane production was only 0.03 L L-1

day-1

(data not shown). This low, non-current driven methane production could be attributed to

the activity of acetoclastic methanogens in the system and/or that syntrophic acetate oxidation

coupled with hydrogenotrophic methanogenesis (Clauwaert and Verstraete 2008). Overall, the

results demonstrate that the current-dependent methane formation was not triggered by the

anode biofilm, but was likely caused by the electron donating cathode.

At cathodic potentials of -1200 and -1600 mV, 82 and 88% of the electron flow in the

RBEC ended up as methane gas (Figure 8.7). While at a cathodic potential of -850 mV (<15

mA), the current could not explain the methane production. This can be due to the fact that the

electrons have by-passed the current generation via alternative methane production pathway(s).

R2

= 0.99R

2= 0.99

0

0.2

0.4

0.60 100 300 400Current (mA)

0

0.5

1.0

1.5Methane Production Rate

COD Removal Rate

(B)

(A)

R2

= 0.92

R2

= 0.98

0

0.2

0.4

2 4 6Electrical Energy Input (kWh kg COD-1)

Me

tha

ne P

rod

uctio

n R

ate

(L L

-1d

ay

-1)

0

0.5

1.0

CO

D R

em

ova

l Ra

te(k

g C

OD

m-3

da

y-1)


- 154 -

Figure 8.7 Electron balance (as electron flow rates) in the methane-producing RBEC with a

cathode poised at -850, -1200 or -1600 mV.

8.3.3 Is Hydrogen a Key End-product of the Cathodic Reduction Reaction in the

Methane-Producing RBEC?

From a thermodynamic perspective, if hydrogen generation (equation 8-4) is the only

cathodic reaction, then for every 10-fold increase in hydrogen partial pressures inside the

headspace the electrode potential is expected to decrease by about 30 mV (Figure 8.8).

Increasing the end product concentration (here H2) would lead to a linear increase of the Gibbs

free energy change in the reaction, rendering the reaction energetically less favorable. On the

contrary, it is expected that by reducing the hydrogen partial pressure in the cathodic headspace,

the cathodic hydrogen production reaction would become more energetically favorable.

However, such phenomenon was not observed in the RBEC reactor. The result shows

that even a 200-fold increase of hydrogen partial pressure in the headspace did not lead to either

a more negative electrode potential or a decrease in current (Figure 8.9). Hence, hydrogen is

unlikely the key cathodic end product that linked the electron transfer from the cathode to

methane production. In the literature, an evidence of methanogenic microorganisms directly

acquire electrons from a cathode to produce methane is recently published (Cheng et al. 2009b).

Further study is required to elucidate the exact mechanism of the cathodic methanogenesis in the

RBEC.

0

50

100

150

200

Electrical

Current

COD

Removal

Hydrogen

Production

Methane

Production

Ele

ctr

on F

low

Ra

te (

mm

ole

L-1

d-1

) -850 mV -1200 mV -1600 mV

0

50

100

150

200

Electrical

Current

COD

Removal

Hydrogen

Production

Methane

Production

Ele

ctr

on F

low

Ra

te (

mm

ole

L-1

d-1

) -850 mV -1200 mV -1600 mV


- 155 -

Figure 8.8 Gibbs free energy change (ΔG°‘) and electrode potential (ΔE°‘) of hydrogen

production reactions (2H+ + 2e

- → H2) at different hydrogen partial pressures.

Values of ΔG°‘ were calculated from (Thauer et al. 1977); ΔE°‘ were calculated

from ΔG°‘ according to ΔG°‘= nFΔE, where n is mole of electron transfer per H2

formed, F is the Faraday‘s constant, 96485 C mol-1

.

Figure 8.9 Current production at different cathodic potentials of the RBEC with either N2

(open symbols) or H2 (solid symbols) in the headspace of the cathode. Note: Prior

to the experiments, the headspace was flushed with pure N2 (1L min-1

) for 5 min.

-100

-80

-60

-40

-20

0

20

40

60

80

100

-30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0

Hydrogen Partial Pressure (Log10 atm)

∆G 0

' (kJ m

ol-1

)

-700

-500

-300

-100

100

300

∆E 0 ' (m

V v

s. A

g/A

gC

l)

Gibbs free energy for H2 production

Electrode Potential for H2 production

0

20

40

60

80

100

-1350-1150-950-750


Cu

rre

nt (m

A)

>99% H2

>99% N2 (~0.5% H2)


- 156 -

8.3.4 Energy Balance

Two approaches were used to calculate the energy recovery from the methanogenic

RBEC process. They are both based on the energy content of the methane recovered, compared

to (1) the electrical energy input plus the energy content in the substrate (i.e. as COD) removed;

and (2) the electrical energy input only (Logan et al. 2008). With the first approach, the energy

recoveries obtained from the three applied cathodic potential settings of -850, -1200 and -1600

mV are 43, 44 and 46%, respectively. The energy content of the organic pollutants in a

wastewater can be considered as a low quality, ―useless‖ energy. Therefore, it is neglected here

in our energy calculation as in the case for conventional anaerobic digestion which usually yield

net energy gain. With this second approach, the energy recoveries obtained from the three

applied cathodic potential settings of -850, -1200 and -1600 mV are 686, 93 and 75%,

respectively.

Figure 8.10 Energy balance (energy recovered as CH4 minus electrical energy input) and COD

removal rate (COD RR) as a function of applied voltage in the methane-producing

RBEC. *Value was re-calculated from this reference as: applied voltage= -825 ±

12 mV; energy balance= -0.14 ± 0.35 kWh kg COD-1

.

The estimated energy balance (energy recovered as CH4 minus electrical energy input) of

the methane-producing RBEC is directly correlated with the applied voltage (Figure 8.10). The

higher the applied voltage, the lower the energy recovery. This result also agrees with Clauwaert

and Verstraete (2008) that a methanogenic bioelectrochemical system may give a negative

y = 0.003x + 2.219

R2 = 0.977

y = -0.001x - 0.097

R2 = 0.993

-2.0

-1.0

0.0

1.0

2.0

-1500-1000-5000

Applied Voltage (mV)

Ou

tpu

t CH

4 -

In

pu

t Ele

c (k

Wh

kg

CO

D-1

)

/ C

OD

RR

(kg

CO

D m

-3 d

ay-1

)

Methane-producing RBEC

Clauwaert and Verstraete (2008)*

COD Removal Rate


- 157 -

energy yield (-0.14 ± 0.35 kWh kg COD-1

). To obtain a positive energy yield one would come at

a cost of a lowered COD removal rate (here 0.2 kg COD m-3

day-1

) (the triangle symbols in

Figure 8.10). Certainly, from a wastewater treatment standpoint this is an undesirable outcome.

Decision needs to be justified on a basis of what the process aims to accomplish.

8.4 Implication of Findings

This study demonstrated that methane rather than hydrogen was the predominant

electron sink of the RBEC operated under anoxic condition. The results are in line with Cheng et

al. (2009) that the methane formation rate was proportional to the cathodic potential and the

applied voltage in a BES. To our current understanding, methane production under reactor

conditions described in the present study can only be of biological but not of abiotic nature. This

implies that both the anodic and cathodic reactions occur in the RBEC were biologically

catalyzed at the electrodes without using chemical catalysts such as platinum. However, whether

the methane producing bacteria can accept electrons directly from the cathode for a one-step

methane formation still remains unclear. Further studies are needed to verify the mechanism of

the bioelectrochemical methane generation and to optimize the RBEC for the proposed

application.

- 158 -

9 Conclusions and Outlook

Chapter Summary

Overall, this thesis has explored the potential of bioelectrochemical systems for energy

recovery from wastewater. It has extended our fundamental understanding on how

electrochemically active microorganisms behave and responsed to the unique environment in

BES. From a practical perspective, the new RBEC configuration may widen the functionality or

suitability of BES for a large-scale wastewater treatment application. However, it remains a

challenge to justify the applicability of BES as a direct one-step wastewater-to-electricity

technology.

This final chapter discusses the potential of bioelectrochemical systems for energy

recovery from wastewater based on the knowledge gained in this thesis. The insights obtained in

this thesis are summarized and discussed. The limitations of the thesis and recommendation for

future research are also addressed.

Chapter 9: Conclusions and Outlook

- 159 -

9.1 Conceptual Progression of the Thesis

9.1.1 A highly Anodophilic Biofilm can be Easily Established from Activated Sludge

This thesis began with evaluating the electricity-producing capacity of an established

electroactive microbial consortium in a classical two-chamber laboratory bioelectrochemical cell

(Chapter 2). An easily obtainable, non-specific mixed culture inoculum (activated sludge from a

wastewater treatment plant) could easily evolve into an anodophilic biofilm on a non-chemically

catalyzed granular graphite anode (bioanode). Similar to most other studies, artificial electron

shuttles (mediators) were not required in the microbial acclimation process. This is a crucial

microbial property because both maintaining sterility and adding artificial electron mediators

(normally they are colored, toxic chemicals) are undesirable requirements for practical BES

operation dealing with wastewater.

Under a well-maintained anodic condition (with the aid of a computer-feedback control),

the established bioanode could effectively generate electricity while removing the organic

pollutant from a wastewater. The established favorable condition included: high buffer strength

and neutral pH in the anolyte (wastewater); efficient mass transfer (provided by a high

recirculation rate of the anolyte to minimize concentration gradient) and an effective ―electron

accepting‖ anode. The low biomass yield (high specific activity) of the bioanode is another merit

of using bioanode for wastewater treatment as it may represent a cost saving in sludge post-

treatment. Nonetheless, the key obstacle of up-scaling (or commercializing) MFC for practical

wastewater treatment application does not lie in the bioanode process.

9.1.1.1 MFC Power Output is undermined by Limitations other than the Metabolic

Capacity of the Bioanode

If we consider only the bioanode capacity and assuming that all other processes in the

MFC (e.g. ionic or ohmic resistance, electrolyte pH and cathodic reduction) are non-limiting, the

potentially retrievable electrical power output is actually more than enough to justify a large-

scale MFC process. According to Rabaey and Verstraete (2005), practical MFC processes aimed

at energy recovery should reach a power output of ≥ 1 kW·m-3

(as compared to conventional

anaerobic digestion technology). This target level has been achieved in Chapter 2 using the

computer-controlled, ferricyanide-cathode MFC in which the capacity of the anodophilic biofilm

would be fully exploited.


- 160 -

Let us consider only the capacity of the established bioanode to anticipate power output

and COD removal in an up-scaled (25 m3) MFC reactor (Figure 9.1). We assume that

technologies could be developed that overcome the generic limitations such as the poor cathodic

oxygen reduction and other electrochemical losses of the process (Box 9.1).

Figure 9.1 Stacking up individual microbial fuel cells in-series to amplify voltage and power

output in a hypothetical up-scaled system. The diagram is not drawn to scale.

Although the estimated COD removal rate is significantly lower compared to the highest

COD removal rate (45 kg COD m-3

d-1

) of the best anaerobic digester available for today

(Expand Granular Sludge Bed) (van Lier et al. 2008), it is well beyond (3-fold) the benchmark

level of 10 kg COD m-3

d-1

for practical MFC up-scaling (Clauwaert et al. 2008a; Pham et al.

2009; Rabaey and Verstraete 2005). Nonetheless, such a promising performance has never been

achieved under realistic condition as very often the capacity of a bioanode is undermined by

other generic limitations such as the poor cathodic oxygen reactivity and the proton

concentration polarization between the two opposing electrodes (pH gradient or pH split).

Addressing these limitations often requires impractical methods. As in Chapter 2, high

performance was obtained by using a ferricyanide-cathode (for effective cathodic reaction), a

high strength phosphate buffer and acid/base dosing (pH control). Although the concept of

polarity inversion invented in Chapter 6 may potentially overcome these generic limitations, it

CO

D

CO

D

CO

D

CO

D

CO

D

CO

D

CO

D

CO

D

CO

D

CO

D

0.1 m

5 m

5 m

2.5 m3

A

n

o

d

i

c

C

h

a

m

b

e

r

Anode

Cathode

Ion exchange membrane

Power Output = 50 kW


- 161 -

remains as a major research challenge to develop a system that can fully exploit the bacterial

capacity to achieve the benchmark target of ≥ 1 kW·m-3

. Perhaps, as already suggested by other

colleagues, the use of BES to recover or even generate other valuable by-products (hydrogen,

methane, 1,3 propanediol, ethanol or butanol…etc), rather than the direct recovery of electricity

from wastewater may have a better chance of commercial success in the technology market

(Rozendal et al. 2008a; Schroder 2008).

Box 9.1 Estimation of the power output and COD removal rate of an up-scaled MFC

wastewater treatment staked module by considering only the bioanode capacity.

In this estimation, we assume that…

10 MFC units are connected in-series as a stack

The bioanode at each unit could sustain an average volumetric current density of

4000A m-3, at an operating cell voltage of 500 mV (anodic potential -400 mV;

cathodic potential +100 mV) (Chapter 2)

100% coulombic efficiency

All other steps in the MFC are not limiting

Power Output:

Power = Voltage × Current

Voltage: Ten MFC units connected in-series gives: 500 × 10 mV = 5 V

Each MFC unit delivers a current of 10000 A (4000 A m-3 × 2.5 m3)

Hence, Power (W) = 5 V × 10000 A = 50 kW

Volumetric power density = 50 kW/ (2.5 × 10) m3

= 2 kW·m-3

COD Removal Rate:

The COD removal rate (here as acetate) can be estimated from the current:

Current delivered by the MFC stack = 4000A m-3 × 25 m3 = 100000A = 360000000 C h-1

= 3731 mol e h-1 (’ 96485; Faraday‘s constant)

Acetate removal rate = 466.4 mol Ac h-1 (÷ 8; 8 e-/ mole Ac)

= 27517 g Ac h-1 (× 59; 59g/ mole Ac)

COD removal rate = 29361 g COD h-1 (× 1.067; 1 g Ac = 1.067 g COD)

= 29.4 kg COD h-1 (÷ 1000; 1000 g/kg)

= 704.7 kg COD d-1 (× 24; 24 h/day)

= 28.2 kg COD m-3

d-1 (÷ 25; 25 m3/stacked MFC)

9.1.1.2 Wastewater Treating-BES requires Innovative R&D Approaches

Looking back to the past decade, the development of MFC technology was very much

inspired ―or influenced‖ by the conventional chemical fuel cell (e.g. PEM H2-FC) domain (e.g.

reactor configurations, sophisticated ion exchange membrane, membrane electrode assembly,


- 162 -

chemical catalysts, etc). To practically implement MFC for wastewater treatment application, we

may need to exploit completely new R&D directions and approaches.

It is worth noting that no matter to what extent we could overcome the limitations in a

MFC, MFC-based energy powering systems (e.g. automobile‘s engines) don‘t compare

favourably with conventional chemical fuel cells (e.g. PEM hydrogen fuel cell). As shown in

Figure 9.2, the current density (volumetric) of a typical MFC is at least four orders of magnitude

lower than that of a typical PEM H2 fuel cell.

Figure 9.2 The power output of conventional hydrogen fuel cell (PEMFC) far exceeds the

power output of microbial fuel cells. (PEMFC data is obtained from Mench et al. (2001);

MFC-Ferricyanide data is from Chapter 2; MFC-Dissolved O2 data is from Clauwaert et

al. (2008a)).

There are several properties of (domestic) wastewaters that may hinder their use as a

feedstock for electricity production in BES:

Low energy content. A large volumetric portion of wastewater consists of purely water

instead of the ―fuel‖ (i.e. organics) (Verstraete et al. 2009): Is that possible to run a

water-less BES fueled with 99% (v/v) organics?

Wastewater is heavy (> 1 kg/L). Distributing or re-circulating wastewater inside the BES

costs energy (mechanical). Taller BES systems may require more of this energy. Yet,

land area availability becomes a concern for a shorter/ flatter system (reactor footprint).

The design of RBEC configuration (Chapter 7) may help minimize the recirculation

requirement. In contrast, H2-driven chemical fuel cells do not face this problem as the

1300

2 0.1

0

200

400

600

800

1000

1200

1400

Po

we

r D

en

sity (

kW

m-3

)

PEMFC MFC-

Ferricyanide

MFC-

Dissolved O2

Not sustainable and practical; For research only

Maybe Sustainable


- 163 -

fuel is in gaseous form, which is light and also has higher energy content (e.g.

compressed H2).

Wastewaters are normally poorly conductive with only low pH buffering capacity. These

explain why by reducing the electrode spacing (i.e. distance between anode and cathode)

or by adding extra chemical buffer into the electrolyte (wastewater), power output of the

BES is enhanced.

9.1.2 Anodophilic Biofilm Affinity for the Anodic Potential

Bioanode is considered as a common component for both MFC and MFC. Hence,

understanding the bacterial behavior of bioanode is of fundamental importance to BES

development. A generic relationship between microbial activity and anodic potential has been

established in Chapter 3. Two parameters, the half-saturation anodic potential (kAP) and the

critical anodic potential (APcrit.) are defined to describe the affinity of an anodophilic biofilm for

the anodic potential. They are expected to aid the quantitative evaluation or modeling of MFC

anodic processes. A similar biofilm-anode affinity concept have been recently proven by others

using both experimental and modeling approaches (Lee et al. 2009; Marcus et al. 2007; Torres et

al. 2007).

From a practical perspective, knowing the affinity of a specific electroactive culture may

be useful for selecting a specific anodic microbial community from a mixed consortium —

―Electrowinning of bacteria” concept (similar to ―Biofilm-plating‖ as suggested by Pham et al.

(2009)). For example, one could selectivity harvest a bacterial culture with a specifically

modified electrode surface by controlling the anode at a specific potential to ―pick up‖ the target

candidate(s) from a microbial consortia. This is a situation similar to the use of a huge

electromagnet (with controllable magnetic strength) to separate magnetized iron materials from

a heap of rubbish. Depending on the purpose of the ―electrowinning‖, the electrode could be first

coated with a unique mediator, enzyme and drug or even preloaded with another microbial

species or consortia.

9.1.3 The Use of Computer Feedback for Dynamic BES Process Control

Computer-based bioprocess control is particularly useful when up-scaling BES processes

for applications such as wastewater treatment. Similar concepts have been widely adopted in the

wastewater treatment industry (e.g. controlled aeration in a sequencing batch reactor (SBR)

based on dissolved oxygen concentration set points). However, in the literature, the use of

computer programs to control BES processes in a dynamic interactive manner is still uncommon.


- 164 -

This thesis has demonstrated the merit of using computer feedback for advanced

operation of a BES process. For example, feedback controlling the external resistance could

dynamically control the electrode potential and power output of a MFC, or even running a cyclic

voltammetry analysis (Chapter 4). Other examples are the alternating supply of substrate and

oxygen to a single biofilm for the polarity inversion study in Chapter 6 and the rotation of

electrode discs of the RBEC in Chapter 7 and 8.

Obviously, the application of computer feedback control deserves further research

advancement. Below are some potential merits it may offer to a large-scale BES wastewater

treatment process:

Process flexibility Different operational regimes can be tailored for specific

needs.

High precision Computer program can synchronize even highly

sophisticated operational demand.

Low maintenance, low

labor cost

BES is predominately controlled by the program; requires

only little attention from the operator.

Easy to gain knowledge on

system performance

Computer can record input (cause) and output (effect)

parameters (involved in the feedback control loop) over time.

Experienced operators can easily establish understanding on

the system for on-going process improvement, or to develop

database of the process.

9.1.4 The Rotatable Bioelectrochemical Contactor: A New Option for Large-Scale

Wastewater Treatment

The discovery that an electrochemically active biofilm could catalyze both anodic and

cathodic reactions in a BES leads to the development of a novel BES configuration, the RBEC.

Two operating modes, MFC and MEC have been tested. By allowing electron flow from the

water-immersed to the air-exposed portion of a biological contactor, organic removal rate can be

accelerated and electrical energy can be directly generated from the process as a MFC. The

described RBEC process has not been optimized yet. Similar to other fixed-film biological


- 165 -

wastewater treatment processes (e.g trickling filter), a significant improvement of both organic

removal and electricity generation is expected by increasing the surface area of the electrode per

reactor volume of the RBEC (e.g. by packing more electrode disc onto the central rotating shaft

without increasing the reactor volume) (Figure 9.3). This may support a higher microbial

activity in the reactor, which may increase the bioelectrochemical conversion of the organics

into electricity. However, this may impose a higher material cost of the process.

Figure 9.3 Hypothetical large-scale RBEC systems for (A) MFC and (B) MEC operations. More

electrode discs are packed inside the reactor to maximize specific surface area for

improved electrochemical reactions.

Cathode

Anode

CODInfluent

CODEffluent

Waste Sludge

Oxygen Oxygen Oxygen

Electricity Output

Recirculation (optional)

Cathode

Anode

CODInfluent

CODEffluent

Waste Sludge

Methane Rich BiogasElectricity Input

Recirculation (optional)

Enclosed Headspace

A.

B.


- 166 -

A possible drawback when scaling up the RBEC is that the ohmic resistance caused by

the ionic flux may become significant in a larger system. This is because the ions (either from

the submerged anode or the exposed cathode) may need to migrate for a longer distance. In

principle, a one-dimensional increase in the disc diameter would lead to a two-dimensional

increase in the disc surface area (Figure 9.4). Although the increased disc diameter would

increase the boundary for ionic flux across the upper and the lower discs, the exponential

increase in the electrode size (i.e. increased current) may potentially render the ionic flux to

become limiting. This concern should be a subject of a further study.

Figure 9.4 Ionic fluxes could be a potential up-scaling limitation of the RBEC: Increasing the

diameter of the RBEC electrode disc leads to exponential increase in electrode surface

area.

Ionic fluxes boundary between anodic and cathodic half discs.


- 167 -

9.1.4.1 RBEC as a MEC for enhanced cathodic oxygen reduction, hydrogen or

methane production

Enhanced Cathodic Oxygen Reduction. The organic removal could be further improved

by adding an extra power supply to the RBEC (up to 1.32 kg COD·m-3

·d-1

). Although net

electrical energy can no longer be obtained, the overall energy consumption is still lower

compared to the conventional activated sludge processes (Table 9.1).

Hydrogen or Methane Production? When the RBEC was operated under fully anaerobic

condition together with the external power supply, hydrogen or methane gas was produced.

Similar to other MECs, the gas production rate in the RBEC was dependent on the applied

power. It has to be noted that in order to produce hydrogen gas as the dominant gas product, a

suitable measure is required to suppress methane formation. The prolonged operation (weeks) of

the RBEC under anaerobic condition resulted in only methane generation. Further studies are

needed to verify the mechanism of the bioelectrochemical methane generation and to optimize

the RBEC for the proposed application.

Table 9.1. Treatment performance and energy comparison of different W.W.T. systems.

Process Status of the Technology Energy Recovery?

COD Removal Rate

Energy Consumption

kg COD·m

-3·d

-1 kWh·kg COD

-1

AS Developed in 1913 and is widely adopted nowadays.

0.5 to 2.0 a 0.7 to 2.0

a

MEC R&D stage (since 2005) as H2 ~ 6.5 a 0.5 to 2.4

a

RBEC (O2 reduction)

No prior art 1.32 0.47

RBEC (H2 production)

No prior art as H2

0.35 b 2.21

RBEC (CH4 production)

No prior art as CH4

0.2 to 1.38 -1.5 to 1.5

AS = activated sludge; MEC = microbial electrolysis cells; a values obtained from Logan et al. (2008)

Methane is a well-known, practical energy carrier in the wastewater industry as it has a

higher volumetric energy density than hydrogen. Using electricity to drive the RBEC process for

methane generation may represent a completely new way to conserve energy surplus from other

alternative renewable energy facilities (e.g. a solar panel or a wind-farm). For example, during a

sunny day where excess electricity is generated by a solar panel a RBEC can convert this energy

into methane while treating a wastewater. The RBEC can be switched back to a MFC-mode for

electricity generation when the energy supply is not available.


- 168 -

At this early stage, particular attention should be given to the optimization of cost

effective materials and the microbial functionality in this novel reactor configuration. Other

considerations such as cost and life-cycle assessments may be required at a later stage to justify

the applicability of the RBEC in a real world situation.

9.2 Final Remarks

Overall, the thesis has explored, from both a fundamental and practical perspectives, the

potential of bioelectrochemical systems for energy recovery from wastewater. Fundamentally,

this thesis has extended our understanding on how electrochemically active microorganisms

behave and response to the unique environment in BES. Especially, the discovery of the

bidirectional microbial electron transfer property may shed light not only on BES development,

but also on the context of fundamental microbiology. Practically, the new RBEC configuration

may widen the functionality or suitability of BES for a large-scale wastewater treatment

application. However, it remains a challenge to justify the applicability of BES as a direct one-

step wastewater-to-electricity technology. Although the noval operational regime of intermittent

prolarity inversion developed in this thesis could be beneficial to the well-known pH gradient

limitation, further process optimization of a practical BES process is still needed. Besides, the

cathodic reaction still remains as the key process bottleneck. Future research efforts should go

towards improving and elucidating in details the mechanisms of the microbe-cathode electron

transfers.

References

- 169 -

R References

Adams, M.W.W., Stiefel, E.I., 1998. Biological hydrogen production: not so elementary.

Science 282(5395), 1842-1843.

Aelterman, P., Freguia, S., Keller, J., Verstraete, W., Rabaey, K., 2008. The anode potential

regulates bacterial activity in microbial fuel cells. Appl Microbiol Biotechnol 78(3), 409-418.

Ajayi, F.F., Kim, K.-Y., Chae, K.-J., Choi, M.-J., 2010. Effect of hydrodymamic force and

prolonged oxygen exposure on the performance of anodic biofilm in microbial electrolysis cells.

International Journal of Hydrogen Energy 35(8), 3206-3213.

Angenent, L., Karim, K., Al-Dahhan, M., Domiguez-Espinosa, R., 2004. Production of

bioenergy and biochemicals from industrial and agricultural wastewater. Trends Biotechnol.

22(9), 477-485.

Bard, A.J., Faulkner, L.R., 2001. Electrochemical methods: fundamentals and applications, 2 ed.

John Wiley & Sons, Inc., New York, USA.

Bard, A.J., Parsons, R., Jordan, J., 1985. Standard potentials in aqueous solution. Marcel Dekker,

New York.

Benemann, J., 1996. Hydrogen biotechnology: Progress and prospects. Nat. Biotechnol. 14,

1101-1103.

Bennetto, H.P., Stirling, J.L., 1983. Anodic reactions in microbial fuel cells. Biotechnol. Bioeng.

25, 559-568.

Biffinger, J., Ribbens, M., Ringeisen, B., Pietron, J., Finkel, S., Nealson, K., 2008.

Characterization of electrochemically active bacteria utilizing a high-throughput voltage-based

screening assay. Biotechnol. Bioeng. DOI:10.1002/bit.22072.

Boeriu, C.G., van Dam, J.E.G., Sanders, J.P.M., 2005. Biomass valorisation for sustainable

development. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for

fuel cells: renewable energy from biomass fermentation, pp. 17-34. IWA Publishing, London,

UK.

Bond, D.R., Lovley, D.R., 2003. Electricity production by Geobacter sulfurreducens attached to

electrodes. Appl. Environ. Microbiol. 69(3), 1548-1555.

Bond, D.R., Lovley, D.R., 2005. Evidence for involvement of an electron shuttle in electricity

generation by Geothrix fermentans. Appl. Environ. Microbiol. 71(4), 2186-2189.

References

- 170 -

Bossel, U., 2006. Does a hydrogen economy make sense? Proc. IEEE 94(10), 1826-1837.

Bryant, M.P., Wolin, E.A., Wolin, M.J., Wolfe, R.S., 1967. Methanobacillus omelianskii, a

symbiotic association of two speices of bacteria. Arch. Microbiol. 59, 20-31.

Call, D., Logan, B.E., 2008. Hydrogen production in a single chamber microbial electrolysis cell

lacking a membrane. Environ. Sci. Technol. 42(9), 3401-3406.

Chachuat, B., Roche, N., Latifi, M.A., 2005. Long-term optimal aeration strategies for small-

size alternating activated sludge treatment plants. Chem. Eng. Process. 44, 593-606.

Chamberlain, J.C., Foley, H.M., MacDonald, G.J., Ruderman, M.A., 1982. Carbon dioxide

review 1982. Oxford University Press, New York, USA.

Chambers, B., Upton, J., Greenhalgh, S.H., 1998. The development and use of hybrid aeration

processes. Wat. Environ. J. 12, 322-329.

Chaudhuri, S.K., Lovley, D.R., 2003. Electricity generation by direct oxidation of glucose in

mediatorless microbial fuel cells. Nat. Biotechnol. 21(10), 1229-1232.

Chen, E., 2003. Thermodynamics and electrochemical kinetics. In: Hoogers, G. (Ed.), Fuel cell

technology handbook. CRC Press LLC.

Cheng, K.Y., Cord-Ruwisch, R., Ho, G., 2007. Limitations of bio-hydrogen production by

anaerobic fermentation process: an overview. American Institute of Physics Conf. Proc. 941,

264-269.

Cheng, K.Y., Cord-Ruwisch, R., Ho, G., 2009a. A new approach for in situ cyclic voltammetry

of a microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry 74(2), 227-231.

Cheng, K.Y., Ho, G., Cord-Ruwisch, R., 2008. Affinity of microbial fuel cell biofilm for the

anodic potential. Environ. Sci. Technol. 42(10), 3828-3834.

Cheng, S., Xing, D., Call, D.F., Logan, B.E., 2009b. Direct biological conversion of electrical

current into methane by electromethanogenesis. Environ. Sci. Technol. 43, 3953-3958.

Choi, Y., Kim, N., Kim, S., Jung, S., 2003. Dynamic behaviors of redox mediators within the

hydrophobic layers as an important factor for effective microbial fuel cell operation. Bull.

Korean Chem. Soc. 24(4), 437-440.

Claassen, P.A.M., van Lier, J.B., Contreras, A.M.L., van Niel, E.W.J., Sijtsma, L., Stams,

A.J.M., de Vries, S.S., Weusthuis, R.A., 1999. Utilisation of biomass for the supply of energy

carriers. Appl. Microbiol. Biotechnol. 52, 741-755.

Clauwaert, P., Aelterman, P., Pham, H.T., De Schamphelaire, L., Carballa, M., Rabaey, K.,

Verstraete, W., 2008a. Minimizing losses in bio-electrochemical systems: the road to

applications. Appl. Microbiol. Biotechnol. 79, 901-913.

References

- 171 -

Clauwaert, P., Mulenga, S., Aelterman, P., Verstraete, W., 2009. Litre-scale microbial fuel cells

operated in a complete loop. Appl. Microbiol. Biotechnol. 83(2), 241-247.

Clauwaert, P., Rabaey, K., Aelterman, P., Schamphelaire, L.D., Pham, T.H., Boeckx, P., Boon,

N., Verstraete, W., 2007a. Biological denitrification in microbial fuel cells. Environ. Sci.

Technol. 41(9), 3354 -3360.

Clauwaert, P., Toledo, R., van der Ha, D., Crab, R., Verstraete, W., Hu, H., Udert, K.M., Rabaey,

K., 2008b. Combining biocatalyzed electrolysis with anaerobic digestion. Wat. Sci. Technol.

57(4), 575-579.

Clauwaert, P., Van Der Ha, D., Boon, N., Verbeken, K., Verhaege, M., Rabaey, K., Verstraete,

W., 2007b. Open air biocathode enables effective electricity generation with microbial fuel cells.

Environ. Sci. Technol. 41(21), 7564-7569.

Clauwaert, P., Verstraete, W., 2008. Methanogenesis in membraneless microbial electrolysis

cells. Appl. Microbiol. Biotechnol. 82(5), 829-836.

Clauwaert, P., Verstraete, W., 2009. Methanogenesis in membraneless microbial electrolysis

cells. Appl. Microbiol. Biotechnol. 82(5), 829-836.

Closed Reflux Colorimetric Method, 1992. Section 5220 D. Standard methods for the

examination of water and wastewater, 18 ed.

Cord-Ruwisch, R., Lovley, D.R., Schink, B., 1998. Growth of Geobacter sulfurreducens with

acetate in syntrophic cooperation with hydrogen-oxidizing anaerobic partners. Appl. Environ.

Microbiol. 64(6), 2232-2236.

Cord-Ruwisch, R., Mercz, T.I., Hoh, C.Y., Strong, G.E., 1997. Dissolved hydrogen

concentration as an on-line control parameter for the automated operation and optimization of

anaerobic digesters. Biotechnol. Bioeng. 56(6), 626-634.

Cord-Ruwisch, R., Seitz, H.-J., Conrad, R., 1988. The capacity of hydrogenotrophic anaerobic

bacteria to compete for traces of hydrogen depends on the redox potential of the terminal

electron acceptor. Arch. Microbiol 149, 350- 357.

Cortez, S., Teixeira, P., Oliveira, R., Mota, M., 2008. Rotating biological contactors: a review on

main factors affecting performance. Rev. Environ. Sci. Biotechnol. 7(2), 155-172.

Davis, J.B., Yarbrough, H.F., 1962. Preliminary experiments on a microbial fuel cell. Science

137, 615-616.

De Graef, M.R., Alexeeva, S., Snoep, J.L., De Mattos, M.J.T., 1999. The steady-state internal

redox state (NADH/NAD) reflects the external redox state and is correlated with catabolic

adaptation in Escherichia coli. J. Bacteriol. 181(8), 2351-2357.

References

- 172 -

de Mes, T.Z.D., Stams, A.J.M., Reith, J.H., Zeeman, G., 2003. Methane production by anaerobic

digestion of wastewater and solid wastes. In: Reith, J.H., Wijffels, R.H., Barten, H. (Eds.), Bio-

methane & Bio-hydrogen: Status and perspectives of biological methane and hydrogen

production, pp. 58-102. Dutch Biological Hydrogen Foundation, The Netherlands.

Di Palma, L., Verdone, N., 2009. The effect of disk rotational speed on oxygen transfer in

rotating biological contactors. Bioresourc. Technol. 100(3), 1467-1470.

Dumas, C., Basseguy, R., Bergel, A., 2008a. Electrochemical activity of Geobacter

sulfurreducens biofilms on stainless steel anodes. Electrochim. Acta 53, 5235-5241.

Dumas, C., Basseguy, R., Bergel, A., 2008b. Microbial electrocatalysis with Geobacter

sulfurreducens biofilm on stainless steel cathodes. Electrochim. Acta 53, 2494-2500.

EIA, 2008. International Energy Outlook 2008. Energy Information Administration, Official

Energy Statistics from the U.S. Government.

Fan, Y., Hu, H., Liu, H., 2007. Sustainable power generation in microbial fuel cells using

bicarbonate buffer and proton transfer mechanisms. Environ. Sci. Technol. 41(23), 8154-8158.

Finkelstein, D.A., Tender, L.M., Zeikus, J.G., 2006. Effect of electrode potential on electrode-

reducing microbiota. Environ. Sci. Technol. 40, 6990-6995.

Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007a. Electron and carbon balances in microbial

fuel cells reveal temporary bacterial storage behavior during electricity generation. Environ. Sci.

Technol. 41(8), 2915-2921.

Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007b. Non-catalyzed cathodic oxygen reduction at

graphite granules in microbial fuel cells. Electrochimica Acta 53(2), 598-603.

Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007c. Sequential anode-cathode configuration

improves cathodic oxygen reduction and effluent quality of microbial fuel cells. Wat. Res. 42,

1387-1396.

Fricke, K., Harnisch, F., Schroder, U., 2008. On the use of cyclic voltammetry for the study of

anodic electron transfer in microbial fuel cells. Energy Environ. Sci. 1, 144-147.

Gil, G.C., Chang, I.S., Kim, B.H., Kim, M., Jang, J.K., Park, H.S., Kim, H.J., 2003. Operational

parameters affecting the performance of a mediator-less microbial fuel cell. Biosensors

Bioelectronics 18, 327-334.

Gregory, K.B., Bond, D.R., Lovley, D.R., 2004. Graphite electrodes as electron donors for

anaerobic respiration. Environ. Microbiol. 6 (6), 596-604.

Hallenbeck, P.C., 2005. Fundamentals of the fermentative production of hydrogen. Water Sci.

Technol. 52, 21-29.

References

- 173 -

Hallenbeck, P.C., Benemann, J.R., 2002. Biological hydrogen production; fundamentals and

limiting processes. Int. J. Hydrogen Energy 27 1185 - 1193.

Hallenbeck, P.C., Ghosh, D., 2009. Advances in fermentative biohydrogen production: the way

forward? Trends in Biotechnol. 27(5), 287-297.

Han, S.K., Shin, H.S., 2004. Biohydrogen production by anaerobic fermentation of food waste.

Int. J. Hydrogen Energy 29 569 - 577.

Harnisch, F., Schroder, U., 2009. Selectivity versus mobility: How to separate anode and

cathode in microbial bioelectrochemical systems? Chem. Sus. Chem. 2(10), 921-926.

Harnisch, F., Schroder, U., Scholz, F., 2008. The suitability of monopolar and bipolar ion

exchange membranes as separators for biological fuel cells. Environ. Sci. Technol. 42(5), 1740-

1746.

He, Z., Angenent, L.T., 2006. Application of bacterial biocathodes in microbial fuel cells.

Electroanalysis 18(19-20), 2009-2015.

He, Z., Minteer, S.D., Angenent, L.T., 2005. Electricity generation from artificial wastewater

using an upflow microbial fuel cell. Environ. Sci. Technol. 39, 5262-5267.

Heyndrickx, M., De Vos, P., De Ley, J., 1991. Fermentation characteristics of Clostridium

pasteurianum LMG 3285 grown on glucose and mannitol. J. Appl. Bacteriol. 70, 52-58.

Hoh, C.Y., 1996. Modelling of reaction kinetics in anaerobic ecosystems with considerations of

thermodynamic principles. School of Biological and Environmental Science. Murdoch

University, Perth, Western Australia.

Hu, H., Fan, Y., Liu, H., 2008. Hydrogen production using single-chamber membrane-free

microbial electrolysis cells. Wat. Res. 42(15), 4172-4178.

Hu, Z., 2008. Electrochemical determination of anaerobic microbial decay coefficients.

Chemosphere 72(2), 312-318.

IEA, 2004. World Energy Outlook 2004. International Energy Agency, Paris, France.

IPCC, 2007. Intergovernmental Panel on Climate Change. Climate Change 2007: Synthesis

Report. United Nations, New York.

Israni, S.H., Koli, S.S., Ratwardhan, A.W., Melo, J.S., D'souza, S.F., 2002. Phenol degradation

in rotating biological contactors. J. Chem. Technol. Biotechnol. 77(9), 1050-1057.

Ito, T., Nakashimada, Y., Senba, K., Matsui, T., Nishio, N., 2005. Hydrogen and ethanol

production from glycerol-containing wastes discharged after biodiesel manufacturing process. J.

Bioscience Bioeng. 100(3), 260-265.

References

- 174 -

Jackson, B.E., McInerney, M.J., 2002. Anaerobic microbial metabolism can proceed close to

thermodynamic limits. Nature 415, 454-456.

Jang, J.K., Pham, T.H., Chang, I.S., Kang, K.H., Moon, H., Cho, K.S., Kim, B.H., 2004.

Construction and operation of a novel mediator- and membrane-less microbial fuel cell. Process

Biochem. 39, 1007-1012.

Jeremiasse, A.W., Hamelers, H.V.M., Buisman, C.J.N., 2010. Microbial electrolysis cell with a

microbial biocathode. Bioelectrochemistry 78(1), 39-43.

Johansson, T.B., Goldemberg, J., 2004. World energy assessment overview: 2004 update.

United Nations Development Programme, UN-DESA and the World Energy Council.

Karube, I., Urano, N., Matsunaga, T., Suzuki, S., 1982. Hydrogen production from glucose by

immobilized growing cells of Clostrium butyricum. Eur. J. Appl. Microbiol. 16, 5-9.

Keighery, P., 2004. Risks in assessing the aeration requirements for an intermittent treatment

plant. Enviro 04 Convention and Exhibition, Sydney, Australia.

Kim, B.H., Chang, I.S., Gadd, G.M., 2007. Challenges in microbial fuel cell development and

operation. Appl Microbiol Biotechnol 76(3), 485-494.

Kim, B.H., Park, H.S., Kim, H.J., Kim, G.T., Chang, I.S., Lee, J., Phung, N.T., 2004a.

Enrichment of microbial community generating electricity using a fuel-cell-type electrochemical

cell. Appl. Microbiol. Biotechnol. 63, 672-681.

Kim, H.J., Park, H.S., Hyun, M.S., Chang, I.S., Kim, M., 2002. A mediator-less microbial fuel

cell using a metal reducing bacterium, Shewanella putrefaciens. Enzyme Microbial Technol. 30,

145-152.

Kim, S.H., Han, S.K., Shin, H.S., 2004b. Feasibility of biohydrogen production by anaerobic co-

digestion of food waste and sewage sludge. Int. J. Hydrogen Energy 29, 1607 - 1616.

Klass, D.L., 1993. Fossil fuel consumption and atmosphric CO2. Energy Policy 21(11), 1076-

1078.

Klass, D.L., 2003. A critical assessment of renewable energy usage in the USA. Energy Policy

31, 353-367.

Kumar, N., Das, D., 2000. Enhancement of hydrogen production by Enterobacter cloacae.

Process Biochem. 35, 589-593.

Lay, J.J., Lee, Y.J., Noike, T., 1999. Feasibility of biological hydrogen production from organic

faction of municipal solid waste. Water Res. 33, 2579-2586.

Lee, C.C., Lin, S.D., 2000. Handbook of Environmental Engineering Calculations, 1 ed.

McGraw-Hill Professional.

References

- 175 -

Lee, H.-S., Torres, C.I., Rittmann, B.E., 2009. Effects of substrate diffusion and anode potential

on kinetic parameters for anode-respiring bacteria. Environ. Sci. Technol. 42, 7571-7577.

Lee, J., Phung, N.T., Chang, I.S., Kim, B.H., Sung, H.C., 2006. Use of acetate for enrichment of

electrochemically active microorganisms and their 16S rDNA analyses. FEMS Microbiol. Lett.

223(2), 185-191.

Lehninger, A.L., Nelson, D.L., Cox, M.M., 1993. Principles of Biochemistry, 2 ed. Worth

Publishers.

Liu, H., 2002. Bio-hydrogen production from carbohydrate-containing wastewater. Department

of Civil Engineering, pp. 1-172. The University of Hong Kong, Hong Kong SAR.

Liu, H., Grot, S., Logan, B.E., 2005. Electrochemically assisted microbial production of

hydrogen from acetate. Environ. Sci. Technol. 39, 4317-4320.

Liu, Y., Harnisch, F., Fricke, K., Sietmann, R., Schroder, U., 2008. Improvement of the anodic

bioelectrocatalytic activity of mixed culture biofilms by a simple consecutive electrochemical

selection procedure. Biosensors. Bioelectronics 24, 1006-1011.

Logan, B., 2005a. Generating electricity from wastewater treatment. Water Environ. Res. , page

211 (editorial).

Logan, B., Regan, J.M., 2006a. Electricity-producing bacterial communities in microbial fuel

cells. Trends Microbiol. 14(12), 512-518.

Logan, B., Regan, J.M., 2006b. Microbial fuel cells: challenges and applications. Environ. Sci.

Technol. 40(18), 5172-5180.

Logan, B.E., 2005b. Simultaneous wastewater treatment and biological electricity generation.

Water Sci. Technol. 52(1-2), 31-37.

Logan, B.E., 2008. Microbial fuel cells. John Wiley & Sons, Inc., Hoboken, New Jersey.

Logan, B.E., Call, D., Cheng, S., Hamelers, H.V.M., Sleutels, T.H.J.A., Jeremiasse, A.W.,

Rozendal, R.A., 2008. Microbial electrolysis cells for high yield hydrogen gas production from

organic matter. Environ. Sci. Technol. 42(23), 8630-8640.

Logan, B.E., Hamelers, B., Rozendal, R.A., Schroder, U., Keller, J., Freguia, S., Aelterman, P.,

Verstraete, W., Rabaey, K., 2006. Microbial fuel cells: methodology and technology. Environ.

Sci. Technol. 40(17), 5181 -5192.

Lovley, D.R., 2006a. Bug juice: harvesting electricity with microorganisms. Nat. Rev. 4, 497-

508.

Lovley, D.R., 2006b. Microbial fuel cells: novel microbial physiologies and engineering

approaches. Curr. Opin. Biotechnol. 17, 327-332.

References

- 176 -

Lovley, D.R., 2008. The microbe electric: conversion of organic matter to electricity. Curr. Opin.

Biotechnol. 19(6), 564-571.

Lovley, D.R., Goodwin, S., 1988. Hydrogen concentrations as an indicator of the predominant

terminal electron-accepting reactions in aquatic sediments. Geochimica et Cosmochimica Acta

52, 2993-3003.

Lovley, D.R., Holmes, D.E., Nevin, K.P., 2004. Dissimilatory Fe(III) and Mn(IV) reduction Adv.

Microb. Physiol. 49 219-286.

Malki, M., De Lacey, A.L., Rodríguez, N., Amils, R., Fernandez, V.M., 2008. Preferential use

of an anode as an electron acceptor by an acidophilic bacterium in the presence of oxygen. Appl.

Environ. Microbiol. 74(14), 4472-4476.

Marcus, K.A., Torres, C.I., Rittmann, B.E., 2007. Conduction-based modelling of the biofilm

anode of a microbial fuel cell. Biotechnol. Bioeng. 98, 1171-1182.

Marken, F., Neudeck, A., Bond, A.M., 2002. Cyclic voltammetry. In: Scholz, F. (Ed.),

Electroanalytical methods: guide to experiments and applications, pp. 51-97. Springer-Verlag,

Berlin Heidelberg.

Mason, C.M., Culvern, J.B., 1949. Electrical conductivity of orthophosphoric acid and of

sodium and potassium dihydrogen phosphates at 25oC. J. Am. Chem. Soc. 71, 2387-2393.

Mench, M.M., Wang, C.-Y., Thynell, S.T., 2001. An introduction to fuel cells and related

transport phenomena. Int. J. Transport Phenomena 3, 151-176.

Menicucci, J., Beyenal, H., Marsili, E., A., V.R., Lewandowski, Z., 2006. Procedure for

determining maximum sustainable power generated by microbial fuel cells. Environ. Sci.

Technol. 40, 1062-1068.

Momirlan, M., Veziroglu, T.N., 2005. The properties of hydrogen as fuel tomorrow in

sustainable energy system for a cleaner planet. Int. J. Hydrogen Energy 30, 795-802.

Myers, C.R., Nealson, K.H., 1990. Respiration-linked proton translocation coupled to anaerobic

reduction of maganese(IV) and iron(III) in Shewanella putrefaciens MR-1. J. Bacteriol. 172,

6232-6238.

Nealson, K.H., Saffarini, D.A., 1994. Iron and manganese in anaerobic respiration:

environmental significance, physiology, and regulation. Annu. Rev. Microbiol. 48, 311-343.

Newman, D.K., Kolter, R., 2000. A role for excreted quinones in extracellular electron transfer.

Nature 405, 94-97.

Odom, J.M., Peck, H.D., 1984. Hydrogenase, electron-transfer proteins, and energy coupling in

the sulfate-reducing bacteria Desulfovibrio. Ann. Rev. Microbiol. 38, 551-592.

References

- 177 -

Oh, S.E., Kim, J.R., Joo, J.H., Logan, B.E., 2009. Effects of applied voltages and dissolved

oxygen on sustained power generation by microbial fuel cells. Wat. Sci. Technol. 60(5), 1311-

1317.

Oh, S.E., Min, B., Logan, B.E., 2004. Cathode performance as a factor in electricity generation

in microbial fuel cells. Environ. Sci. Technol. 38, 4900-4904.

Park, D.H., Kim, S.K., Shin, I.H., Jeong, Y.J., 2000. Electricity production in biofuel cell using

modified graphite electrode with neutral red. Biotechnol. Lett. 22, 1301-1304.

Park, D.H., Zeikus, J.G., 2000. Electricity generation in microbial fuel cells using neutral red as

an electronophore. Appl. Environ. Microbiol. 66(4), 1292-1297.

Parot, S., Delia, M.-L., Bergel, A., 2007. Forming electrochemically active biofilms from garden

compost under chronoamperometry. Bioresour. Technol. 99(11), 4809-4816.

Pham, H.T., Boon, N., Aelterman, P., Clauwaert, P., Schamphelaire, L.D., Vanhaecke, L.,

Maeyer, K.D., Hofte, M., Verstraete, W., Rabaey, K., 2007. Metabolites produced by

Pseudomonas sp. enable a Gram-positive bacterium to achieve extracellular electron transfer.

Appl. Microbiol. Biotechnol. DOI 10.1007/s00253-007-1248-6.

Pham, H.T., Rabaey, K., Aelterman, P., Clauwaert, P., De Schamphelaire, L., Verstraete, W.,

2006. Microbial fuel cells in relation to conventional anaerobic digestion technology. Eng. Life

Sci. 6(3), 285-292.

Pham, T.H., Aelterman, P., Verstraete, W., 2009. Bioanode performance in bioelectrochemical

systems: recent improvements and prospects. Trends Biotechnol. 27(3), 168-178.

Picioreanu, C., Loosdrecht, M.C.M., Cirtis, T.P., Scott, K., 2009. Model based evaluation of the

effect of pH and electrode geometry on microbial fuel cell performance. Bioelectrochemistry

doi:10.1016/j.bioelechem.2009.04.009.

Potter, M.C., 1911. Electrical effects accompanying the decomposition of organic compounds.

Proc. Roy. Soc. London, Series B. 84, 260-276.

Qiao, Y., Li, C.M., Bao, S.-J., Lu, Z., Hong, Y., 2008. Direct electrochemistry and

electrocatalytic mechanism of evolved Escherichia coli cells in microbial fuel cells. Chem.

Commun., 1290-1292.

Rabaey, K., Boon, N., Hofte, M., Verstraete, w., 2005a. Microbial phenazine production

enhances electron transfer in biofuel cells. Environ. Sci. Technol. 39, 3401-3408.

Rabaey, K., Boon, N., Siciliano, S.D., Verhaege, M., Verstraete, W., 2004. Biofuel cells select

for microbial consortia that self-mediate electron transfer. Appl. Environ. Microbiol. 70(9),

5373-5382.

References

- 178 -

Rabaey, K., Clauwaert, P., Aelterman, P., Verstraete, W., 2005b. Tubular microbial fuel cells for

efficient electricity generation. Environ. Sci. Technol. 39, 8077-8082.

Rabaey, K., Lissens, G., Siciliano, S.D., Verstraete, W., 2003. A microbial fuel cell capable of

converting glucose to electricity at high rate and efficiency. Biotechnol. Lett. 25, 1531-1535.

Rabaey, K., Lissens, G., Verstraete, w., 2005c. Microbial fuel cells: performances and

perspectives. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for fuel

cells: renewable energy from biomass fermentation, pp. 375-396. IWA Publishing, London, UK.

Rabaey, K., Read, S.T., Clauwaert, P., Freguia, S., Bond, P.L., Blackall, L.L., Keller, J., 2008.

Cathodic oxygen reduction catalyzed by bacteria in microbial fuel cells. ISME J., 1-9.

Rabaey, K., Rodrı´guez, J., Blackall, L.L., Keller, J., Gross, P., Batstone, D., Verstraete, W.,

Nealson, K.H., 2007. Microbial ecology meets electrochemistry: electricity-driven and driving

communities. ISME J. 1, 9-18.

Rabaey, K., Verstraete, W., 2005. Microbial fuel cells: novel biotechnology for energy

generation. Trends Biotechnol. 23(6), 291-298.

Razi, F., Yassin, A.A.A., Lyuke, S.E., Ngan, M.A., Morimoto, M., 2005. Bio-hydrogen

systhesis from wastewater by anaerobic fermentation using microflora. Int. J. Green Energy 2,

387-396.

Reguera, G., D., M.K., Mehta, T., Nicoll, J.S., Tuominen, M.T., Lovley, D.R., 2005.

Extracellular electron transfer via microbial nanowires. Nature 435, 1098-1101.

Reguera, G., Nevin, K.P., Nicoll, J.S., Covalla, S.F., Woodard, T.L., Lovley, D.R., 2006.

Biofilm and nanowire production leads to increased current in Geobacter sulfurreducens fuel

cells. Appl. Environ. Microbiol. 72, 7345-7348.

Richter, H., Nevin, K.P., Jia, H., Lowy, D.A., Lovley, D.R., Tender, L.M., 2009. Cyclic

voltammetry of biofilms of wild type and mutant Geobacter sulfurreducens on fuel cell anodes

indicates possible roles of OmcB, OmcZ, type IV pili, and protons in extracellular electron

transfer. Energy Environ. Sci. 2, 506-516.

Rifkin, J., 2002. The Hydrogen Economy. Tarcher.

Rismani-Yazdi, H., Carver, S.M., Christy, A.D., Tuovinen, O.H., 2008. Cathodic limitation in

microbial fuel cells: an overview. J. Power Sources 180(2), 683-694.

Rittmann, B.E., 2008. Opportunities for renewable bioenergy using microorganisms. Biotechnol.

Bioeng. 100(2), 203-212.

Rozendal, R.A., 2007. Hydrogen production throught biocatalyzed electrolysis. Wageningen

University, Wageningen, the Netherlands.

References

- 179 -

Rozendal, R.A., Hamelers, H.V.M., Buisman, C.J.N., 2006a. Effects of membrane cation

transport on pH and microbial fuel cell performance. Environ. Sci. Technol. 40(17), 5206-5211.

Rozendal, R.A., Hamelers, H.V.M., Euverink, G.J.W., Metz, S.J., Buisman, C.J.N., 2006b.

Principle and perspectives of hydrogen production through biocatalyzed electrolysis. Int. J.

Hydrogen Energy 31, 1632-1640.

Rozendal, R.A., Hamelers, H.V.M., Rabaey, K., Keller, J., Buisman, C.J.N., 2008a. Towards

practical implementation of bioelectrochemical wastewater treatment. Trends Biotechnol. 26(8),

450-459.

Rozendal, R.A., Jeremiasse, A.W., Hamelers, H.V.M., Buisman, C.J.N., 2008b. Hydrogen

production with a microbial biocathode. Environ. Sci. Technol. 42(2), 629-634.

Ruzicka, M., 1996. The effect of hydrogen on acidogenic glucose cleavage. Water Res. 30(10),

2447-2451.

Schink, B., 1997. Energetics of syntrophic cooperation in methanogenic degradation. Microbiol.

Mol. Biol. Rev. 61, 262-280.

Schnurer, A., Schink, B., Svensson, B.H., 1996. Clostridium ultunense sp. nov., a mesophilic

bacterium oxidizing acetate in syntrophic association with a hydrogenotrophic methanogenic

bacterium. Int. J. Syst. Bacteriol. 46(4), 1145-1152.

Scholz, F., 2002. Electroanalytical methods: guide to experiments and applications. Springer-

Verlag Berlin Heidelberg.

Schroder, U., 2007. Anodic electron transfer mechanisms in microbial fuel cells and their energy

efficiency. Phys. Chem. Chem. Phys. 9, 2619-2629.

Schroder, U., 2008. From wastewater to hydrogen: biorefineries based on microbial fuel-cell

technology. Chem. Sus. Chem. 1(4), 281-282.

Sears, H.J., Spiro, S., Richardson, D.J., 1997. Effect of carbon substrate and aeration on nitrate

reduction and expression of the periplasmic and membrane-bound nitrate reductases in carbon-

limited continuous cultures of Paracoccus denitrificans Pd1222. Microbiology 143, 3767-3774.

Stams, A.J.M., de Bok, F.A.M., Plugge, C.M., van Eekert, M.H.A., Dolfing, J., Schraa, G., 2006.

Exocellular electron transfer in anaerobic microbial communities. Environ. Microbiol. 8(3), 371-

382.

Sund, C.J., McMasters, S., Crittenden, S.R., Harrell, L.E., Sumner, J.J., 2007. Effect of electron

mediators on current generation and fermentation in a microbial fuel cell. Appl Microbiol

Biotechnol 76(3), 561-568.

References

- 180 -

Svensson, B.H., Karlsson, A., 2005. Dark fermentation for hydrogen production from organic

wastes. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for fuel cells:

renewable energy from biomass fermentation, pp. 209-220. IWA Publishing, London, UK.

Taguchi, F., Chang, J.D., Mizukami, N., Saito-Taki, T., Hasegawa, K., Morimoto, M., 1993.

Isolation of a hydrogen-producing bacterium, Clostridium beijerinckii strain AM21B, from

termites. Can. J. Microbiol. 39, 726-730.

Tanisho, S., 2001. A scheme for developing the yield of hydrogen by fermentation. In: Miyake,

J., Matsunnaga, T., San Pietro, A. (Eds.), Biohydrogen II, pp. 131-138. Elsevier Science,

Amsterdam.

Tanisho, S., Kamiya, N., Wakao, N., 1989. Microbial fuel cell using Enterobacter aerogenes.

Bioelectrochem. Bioenergetics 21, 25-32.

Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater engineering, treatment and

reuse, 4 ed. McGraw-Hill, Boston.

Thauer, R.K., Jungermann, K., Decker, K., 1977. Energy conservation in chemotrophic

anaerobic bacteria. Bacteriol. Rev. 41(1), 100-180.

Torres, C.I., Marcus, A.K., Parameswaran, P., Rittmann, B.E., 2008a. Kinetic experiments for

evaluating the Nernst-Monod model for anode-respiring bacteria (ARB) in a biofilm anode.

Environ. Sci. Technol. 42(17), 6593-6597.

Torres, C.I., Marcus, A.K., Rittmann, B.E., 2007. Kinetics of consumption of fermentation

products by anode-respiring bacteria. Appl. Microbiol. Biotechnol. DOI 10.1007/s00253-007-

1198-z.

Torres, C.I., Marcus, A.K., Rittmann, B.E., 2008b. Proton transport inside the biofilm limits

electrical current generation by anode-respiring bacteria. Biotechnol. Bioeng. 100, 872-881.

Van Aalst-van Leeuwen, M.A., Pot, M.A., Van Loosdrecht, M.C.M., 1997. Kinetic modelling of

poly(β-hydroxybutyrate) production and consumption by Paracoccus pantotrophus under

dynamic substrate supply. Biotechnol. & Bioeng. 55(5), 773-782.

van Ginkel, S.W., Logan, B., 2005. Increased biological hydrogen production with reduced

organic loading. Water Res. 39, 3819-3826.

van Lier, J.B., Mahmoud, N., Zeeman, G., 2008. Anaerobic wastewater treatment. In: Henze, M.,

van Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D. (Eds.), Biological Wastewater Treatment:

Principles, Modelling and Design. IWA Publishing, London, UK.

van Lier, J.B., Tilche, A., Ahring, B.K., Macarie, H., Moletta, R., Dohanyos, M., Pol, L.W.H.,

Lens, P., Verstraete, W., 2001. New perspectives in anaerobic digestion. Water Sci. Technol. 43,

1-18.

References

- 181 -

Van Loosdrecht, M.C.M., Pot, M.A., Heijnen, J.J., 1997. Importance of bacterial storage

polymers in bioprocesses. Wat. Sci. Technol. 35, 41-47.

Van Spanning, R.J.M., de Boer, A.P.N., Reijnders, W.N.M., De Gier, J.-W.L., Delorme, C.O.,

Stouthamer, A.H., Westerhoff, H.V., Harms, N., van der Oost, J., 1995. Regulation of oxidative

phosphorylation: The flexible respiratory network of Paracoccus denitrificans. J. Bioenergetics

Biomembranes.

Verstraete, W., Van de Caveye, P., Diamantis, V., 2009. Maximum use of resources present in

domestic "used water". Bioresourc. Technol. 100, 5537-5545.

Verstraete, W., van Vaerenbergh, E., 1986. Aerobic activated sludge. In: Rehm, H.J., Reed, G.

(Eds.), Biotechnology, pp. 43-112. VCH, Weinheim, Germany.

von Sperling, M., de Lemos Chernicharo, C.A., 2005. Biological wastewater treatment in warm

climate regions. IWA Publishing, London, UK.

Vrije, T.d., Claassen, P.A.M., 2003. Dark hydrogen fermentations. In: Reith, J.H., Wijffels, R.H.,

Barten, H. (Eds.), Bio-methane & Bio-hydrogen: status and perspectives of biological methane

and hydrogen production, pp. 103-123. Dutch Biological Hydrogen Foundation, The Hague,

Netherlands.

Wang, B., Han, J.-I., 2008. A single chamber stackable microbial fuel cell with air cathode.

Biotechnol. Lett. 31(3), 1573-6776.

Wang, Y.F., Tsujimura, S., Cheng, S.S., Kano, K., 2007. Self-excreted mediator from

Escherichia coli K-12 for electron transfer to carbon electrodes. Appl. Microbiol. Biotechnol. 76,

1439-1446.

Watanabe, K., 2008. Recent developments in microbial fuel cell technologies for sustainable

bioenergy. J. Biosci. Bioeng. 106(6), 528-536.

Yoon, S.-M., Choi, C.-H., Kim, M., Hyun, M.S., Shin, S.-H., Yi, D.-H., Kim, H.J., 2007.

Enrichment of electrochemically active bacteria using a three-electrode electrochemical cell. J.

Microbiol. Biotechnol. 17(1), 110-115.

You, S., Zhao, Q., Zhang, J., Jiang, J., Wan, C., Du, M., Zhao, S., 2007. A graphite-granule

membrane-less tubular air-cathode microbial fuel cell for power generation under continuously

operational conditions. J. Power Sources 173(1), 172-177.

You, S., Zhao, Q., Zhang, J., Jiang, J., Zhao, S., 2006. A microbial fuel cell using permanganate

as the cathodic electron acceptor. J. Power Sources 162, 1409-1415.

Yu, E.H., Cheng, S., Scott, K., Logan, B., 2007. Microbial fuel cell performance with non-Pt

cathode catalysts. J. Power Sources 171, 275-281.

References

- 182 -

Zhang, T., Liu, H., Fang, H.H.P., 2003. Biohydrogen production from starch in wastewater

under thermophilic condition. J. Environ. Manage. 69, 149-156.

Zhao, F., Harnisch, F., Schroder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2005. Application

of pyrolysed iron (II) phthalocyanine and CoTMPP based oxygen reduction catalysts as cathode

materials in microbial fuel cells. Electrochem. Commun. 7, 1405-1410.

Zhao, F., Harnisch, F., Schroder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2006a. Challenges

and constraints of using oxygen cathodes in microbial fuel cells. Environ. Sci. Technol. 40(17),

5193-5199.

Zhao, F., Harnisch, F., Schroeder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2006b.

Challenges and constraints of using oxygen cathodes in microbial fuel cells. Environ. Sci.

Technol. 40(17), 5193-5199.

- 183 -

A Appendixes

Appendix 1 (Supplementary Note to Chapter 1, Section 1.1.2.2.2)

Limitations of Fermentative Hydrogen Production: An Overview

(This article has been published in American Institute of Physics Conf. Proc. 941, 264-269)

Abstract

Turning organic wastes into hydrogen (H2) by using anaerobic fermentative technology can

accomplish both waste treatment and energy recovery objectives. However, H2 production using

anaerobic fermentative approaches faces a fundamental limitation of poor H2 production yield

(i.e. < 4 mol H2 per mol of hexose). This article gives an overview on the limitations of bio-H2

production using fermentative processes. Fundamental microbiology and thermodynamic of the

processes are discussed.

1. Introduction

By decoupling H2 production from methane production, conversion of organic matter

into H2 could be achieved (Vrije and Claassen 2003). Many types of organic compounds,

ranging from polymers to monomers such as carbohydrates, fats and amino acids are known to

be the substrates for H2 production (Claassen et al. 1999). As such, organic wastes used for the

production of methane can also be the potential substrates for anaerobic fermentative production

of H2 (Svensson and Karlsson 2005). In fact, many studies indicate that various wastes

containing high organic matter have been used to produce H2 by anaerobic fermentation process

Appendix 1

- 184 -

(Angenent et al. 2004; Han and Shin 2004; Kim et al. 2004b; Lay et al. 1999; Razi et al. 2005;

van Ginkel and Logan 2005; Zhang et al. 2003). The major differences between the two

processes are that successful biological H2 production requires inhibition of H2-consuming

microorganisms and maximization of H2 production yield from the organic substrates. Most of

the studies revealed that only 10-20% of stoichiometric maximum yield (i.e. 12 moles of H2 per

mole hexose) of H2 could be recovered in anaerobic fermentative processes (Benemann 1996;

Hallenbeck 2005; Hallenbeck and Benemann 2002). Hence, some researchers even argued that

fermentative H2 production should only be restricted to a pre-treatment step in a larger bioenergy

production concept (Angenent et al. 2004). The present article will discuss, from a

microbiological and thermodynamic perspective, the limitations of anaerobic fermentation as H2

production process.

2. Biochemical Pathways of Fermentative Hydrogen Production

Among the many species of anaerobic fermentative bacteria capable of producing H2, the

H2-producing characteristics of two genera, Clostridium (e.g. C. pasteurianum (Heyndrickx et al.

1991); C. beijerinckii (Taguchi et al. 1993) and C. butyricum (Karube et al. 1982) and

Enterobacter (e.g. E. aerogenes (Tanisho et al. 1989) and E. Cloacae (Kumar and Das 2000),

have been studied extensively. Figure 1 illustrates the pathway of anaerobic fermentation by

using glucose as model substrate (Tanisho 2001). The glucose monomer produced from

hydrolysis is taken up by the fermentative bacteria and degraded predominantly through the

Embden-Meyerhof-Parnas (glycolysis) pathways to generate ATP from ADP, leading to the

formation of pyruvate (CH3COCOOH). Like other bacteria or higher eukaryotic cells, the

glycolytic reactions in glucose fermentative bacteria also produce electron equivalents in the

form of NADH, which has to be re-oxidized in order to continue substrate degradation.

Inorganic electron acceptors are usually the preferred candidate in anaerobic respiration for the

bacteria to regenerate these reducing equivalents because they have higher redox potential.

While in the absence of external inorganic electron acceptors, NADH is commonly reoxidized

by H+ and produce H2 and NAD

+. The pyruvate produced from glycolysis is then converted to

acetyl-CoA (CH3COSCoA), liberating carbon dioxide and H2. The pyruvate may also be

converted into acetyl-CoA and formate (CHOOH), which may be converted to H2 and carbon

dioxide by fermentative bacteria such as Escherichia coli. Depending on the microorganisms

Appendix 1

- 185 -

and the environment, the Acetyl-CoA is eventually converted into acetate (CH3COOH), butyrate

(CH3CH2CH2COOH) or ethanol (CH3CH2OH).

Figure A1.1. Pathways of fermentation of glucose. Adapted from (Liu 2002).

The formation of different end products from the pyruvate is found to be highly

dependent on the H2 partial pressure of the system. For instance, since the conversion of glucose

to butyrate, CO2 and H2 yield only 3 mol of ATP per mol of glucose, whereas the conversion of

glucose to acetate, CO2 and H2 can yield 4 mol of ATP per mol of glucose. The production of H2

and acetate from pyruvate is therefore usually preferred by most bacteria with hydrogenase as it

allows the bacteria to conserve more energy from their substrates (Thauer et al. 1977). In fact,

from bio-H2 production stand point, formation of acetate as the end product is also preferred

because it allows the greatest possible amount of H2 to be produced via fermentative pathway.

However, such preferred conversion is only possible when the H2 concentration (or

NADH/NAD+ ratio) in the system is kept low. Under elevated H2 concentration environment,

which may be due to accumulation of H2 because of ineffective H2 removal (e.g. inhibition of

H2-consuming methanogen), H2 production from NADH can be inhibited due to thermodynamic

C6H12O6

2 ADP

2 ATP

2 NAD+

2 NADH + 2 H+

2 H2

2 CH3COCOOH

2 CH3COSCoA

CH3CH2CH2COOH

2 CH3COOH2 CH3CH2OH

4 NAD+ 4 NADH + 4 H+ ADP ATP

2 CO22 H2

2 CHOOH

2 H2

2 CO2

ADP

ATP 2 NAD+

2 NADH + 2 H+

Appendix 1

- 186 -

reasons, leading to the production of propionate or butyrate from pyruvate as alternative electron

sinks (Figure A1.2).

3. Elimination of Syntrophic Partners in Fermentative Hydrogen Production

In methanogensis process, complex organic matter is degraded in a sequence of reactions

by several distinct groups of microorganisms. Fermentative microorganisms will first

breakdown the organic matter into simpler substance, such as H2, formate and acetate, which

serve as the substrates for the methanogenic microorganisms. Also, a variety of other organic

compounds like lactate, ethanol, propionate, butyrate, etc. are formed, and are degraded by

proton-reducing acetogenic bacteria to form the methanogenic substrates, i.e. H2 and acetate.

However, these organisms can only grow and keep converting the organic matter into the

methanogenic substrates when the H2 concentration is maintained at low level by their

syntrophic partner, i.e. methanogenic archaea (Bryant et al. 1967; Schink 1997; Schnurer et al.

1996). Such H2 (or electron + proton) transfer mechanism is an integral and vital process in the

anaerobic mineralization of organic matter, and is commonly regarded as interspecies

interactions between fermentative H2-producing organisms and H2-consuming methanogen

(Odom and Peck 1984).

From a bioenergetic perspective, when organic electron acceptors such as pyruvate or

fumurate were used, the partial oxidation of the substrates results in the formation of products

that still possess high free energy. This explains why anaerobic catabolic reactions involving

organic electron acceptors usually have low free energy change (∆G), and only little energy

could be conserved from the substrate metabolism by the microorganisms (Thauer et al. 1977).

Therefore, the reactions proceed very close to the dynamic equilibrium, and require constant

removal of the end-products of the reactions before the reactions can proceed further (Jackson

and McInerney 2002; Schink 1997). Therefore, the presence of methanogenic organisms is vital

because they can effectively and indirectly maintain the continuous fermentative degradation of

Appendix 1

- 187 -

At Hydrogen Partial

Pressure lowers than

10 Pa:

At Hydrogen Partial

Pressure higher than

10 Pa:

Figure A1.2. Formation of acetate, propionate or butyrate via different pathways of

glucose/pyruvate degradation at elevated or low hydrogen concentrations. H2 and acetate

are produced at low H2 concentration (< 10 Pa) whilst propionate and butyrate are

produced as alternative electron sinks at elevated H2 due to thermodynamic reasons.

Adapted from (Hoh 1996).

organic matter into their own substrates, including H2. However, in fermentative H2 production

process, growth of methanogens is considered as an undesirable process and should be inhibited

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP ATP

Acetate AcetateH2H2

H2H2

CO2 CO2

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP ATP

Acetate AcetateH2H2H2H2

H2H2H2H2

CO2 CO2

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP

Acetate Propionate

H2

CO2

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP

Acetate Propionate

H2H2

CO2

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP

Acetate

Butyrate

H2

CO2

H2

CO2

Glucose

Pyruvate Pyruvate

ATP ATP

NADH NADH

ATP

Acetate

Butyrate

H2H2

CO2

H2H2

CO2

Appendix 1

- 188 -

in order to prevent any consumption of H2, allowing the H2 produced from fermentative

degradation of organic matter by fermentative or acetogenic bacteria to be recovered.

4. Conversion of Fermentation End-Products into Hydrogen: A Thermodynamic

Consideration

A variety of fermentative end-products (mainly as VFAs) are remained after the

fermentative process (refer to Figure A1.1). However, a considerable amount of chemical energy

is still remaining ―unexplored‖ from the process. In methanogenesis, degradation of VFAs such

as acetate is only possible via interspecies H2 transfer, when H2 is kept to a low concentration by

the H2-consuming methanogens, accumulation of H2 to a certain level can result in the inhibition

of VFA degradation reactions due to thermodynamic limitations (Thauer et al. 1977). This

explains why dissolved H2 concentration is considered as a key controlling factor in the

anaerobic processes and it can regulate the rates at which the VFAs degrade to acetate and H2

(Cord-Ruwisch et al. 1997; Lovley and Goodwin 1988). The standard Gibbs free energy change

of reaction (∆Go) is the amount of energy being conserved when the substrate(s) are being

converted into products (s) (Thauer et al. 1977). It can be calculated from the Gibbs free energy

changes of formation from the elements (∆Go

f) of the substrates and products according to the

following equation:

∆Go =

∑

∆G

of of products - ∑

∆G

of of substrates

Since almost all reactions occur in biological systems are not under standard condition (i.e. 298

K, 1 atm pressure, 1 M concentration for all species). The Gibbs free energy change of reaction

corrected for the actual reacting conditions (∆G) can then be calculated from the ∆Go and the

concentrations of all reactants (S) and products (P) according to the following equation:

∆G = ∆G

o + RT log [P]/[S]

Where, R = universal gas constant (8.314 x 10-3

kJ∙mol-1

∙K-1

); T = absolute temperature (K).

(Negative values of ∆G indicate the forward reaction is exergonic or spontaneous,

Appendix 1

- 189 -

and positive values of ∆G indicate the backward reaction is exergonic or

spontaneous.)

According to the literature, amongst all of the end-products commonly formed during

fermentative H2 production, acetate is considered as the key end-product because further

conversion of acetate into H2 is unfavorable to occur in fermentative processes. Based on Gibbs

free energy calculations, H2 production from acetate requires a net energy input of 104.6 kJ mol

acetate-1

under standard conditions (Thauer et al. 1977).

Figure A1.3. Relationship between partial pressure of hydrogen (PH2) and change-in-free-

energy (∆G) for the following hydrogen-producing reactions:

∆Go‘

(kJ/reaction) PH2, ∆G=0

(A) Acetate- + 4H2O → 2HCO3

- + 4H2 + H

+ +104.6 -4.60

(B) Propionate- + 3H2O → acetate

- + 1HCO3

- + 3H2 + H

+ +76.5 -4.49

(C) Butyrate- + 2H2O → acetate

- + 2H2 + H

+ +48.3 -4.25

(D) Ethanol + H2O → acetate- + 2H2 + H

+ +9.6 -0.85

(E) Glycerol + 2H2O → acetate- + HCO3

- + 3H2 + 2H

+ -84.6 +4.96

-10

-5

0

5

10

15

-100-80-60-40-20020406080100120

-10

-5

0

5

10

15

-100-80-60-40-20020406080100120

PH2

(Log10 atm)

∆G at pH 7, 25oC

(A)

(B)

(C)

(D)

(F)

(E)

Appendix 1

- 190 -

(F) Glucose + 4H2O → 2 acetate- + 2HCO3

- + 4H2 + 4H

+ -206.1 +9.06

Notes: PH2, ∆G=0 = hydrogen partial pressure (Log10 atm) when ∆G is equal to zero; ∆G

values (at pH 7 and 25oC) were calculated from the standard Gibbs free energy change (∆G

o‘)

value in (Thauer et al. 1977) using equation ∆G = ∆Go‘

+ 2.47 x ln ([products]/[substrates]).

Concentrations used for all compounds were assumed to be 1M.

Figure A1.3 illustrates the relationship between H2 partial pressure and change-in-free-

energy for several H2 producing reactions that possibly occur in fermentative process. Under

standard condition (i.e. 298 K, 1 atm pressure, 1 M concentration for all species), only the

oxidation reactions of 1 mol glucose to 4 mol H2 and 1 mol glycerol to 3 mol H2 have negative

∆Go‘

values, while all the other reactions have positive ∆Go‘

values, meaning that except for

glucose and glycerol, all the other reactions are energetically unfavorable to occur under

standard condition.

Since ∆G value of each specific reaction is in a linear function of partial pressure of the

product H2, the reaction may proceed toward the product side when the partial pressure of H2

has reduced to a certain level. When ∆G is equal to zero, oxidation of 1 mol acetate to 4 mol H2

only become energetically favorable when the partial pressure of H2 in the system is lower than -

4.6 Log10 atm (i.e. 2.55 Pa), which is extremely low that may even not allow the synthrophic

action of the hydrogenotrophic methanogenic bacteria in anaerobic fermentative system to occur.

According to Figure A1.3, glycerol should also be considered as a promising feedstock to

produce H2 via fermentative processes. This is because the conversion of glycerol into H2 is

energetically favorable (i.e. ∆Go‘

= -84.6 kJ/reaction), and the reaction is only susceptible to the

inhibitory effect of H2 accumulation with H2 partial pressure greater than +4.96 Log10 atm (i.e.

9.24 x 109 Pa), which is very high and is not supposed to happen under biological conditions. In

fact, glycerol is the major by-product of bio-diesel generation processes, and a further increase

in the production of bio-diesel fuels would raise the problem of efficiently treating wastes

containing glycerol (Ito et al. 2005). Therefore, apart from acetate, the possibility of using

glycerol as the feedstock substrates in H2 production process also worthwhile to be investigated.

As aforementioned, even the fermentative process can be successfully manipulated to use

acetate as the substrate for the production of H2, the produced H2 would be under very low

partial pressure, i.e. concentration. This is impractical to obtain large quantity of H2 using

Appendix 1

- 191 -

acetate as the only substrate from fermentative process. In order for the microbes to ―extract‖ the

electrons from the substrate in a more effective manner, it is reasonable to think of using some

approaches to remove the H2 that is produced from the bacteria immediately so as to maintain a

low H2 partial pressure for the H2 producing reactions to remain energetically favorable.

In methanogenesis process where methane forming consortia are present, effective H2

removal is normally being achieved by the methanogenic partners, which are in syntrophic

relationship at close proximity with the H2-producing bacteria. As H2 is one of the substrates for

methanogens, H2 is taken up by them immediately and effectively to produce methane, which

will then be easily collected in gaseous form due to their poor aqueous solubility. In other words,

the distance between the locations of H2 production and H2 consumption is so close to enable

effective mass transfer of H2 between the two partners. In fact, effective mass transfer of H2

between H2-producing and –consuming bacteria can be revealed by estimating the time required

for H2 removal from H2-producing bacteria in a methanogenic environment where H2 is

maintained at low level by syntrophic microbial association.

Time required for the H2-producing bacteria to produce H2 until Inhibitory Levels are reached

under Methanogenic Conditions:

Assume: Methane production rate = 1 L-1

CH4 L-1

day-1

Since: 1 mol of methane formed required 4 mol of H2, i.e. 4H2 + CO2 ↔ CH4 + 2H2O; and

molar volume of CH4 and H2 are 24.5 L mol-1

at 25 oC, 1 atm.

Then: H2 production rate = 4 L-1

H2 L-1

day-1

= 0.16 mol H2 L-1

day-1

= 0.16 M day-1

= 1.85 x 10-6

M second-1

= 1.85 µM second-1

Assume: Saturation concentration of H2 in water (25 oC,1 atm) is 1 mmol L

-1 (Ruzicka 1996)

Appendix 1

- 192 -

If: H2 production was inhibited at PH2 of 10 Pa (Cord-Ruwisch et al. 1998); i.e. 1 x 10-4

mmol L-1

= 100 nM = 0.1 µM

Then: Time Required to Produce H2 until Inhibitory Levels are reached would be:

= 0.1 µM/ 1.85 µM second-1

= 54 milliseconds

This simple calculation suggests that the methanogenic partner (H2-consuming

microorganisms) could effectively take up the H2 produced by the H2-producing bacteria within

a very short time in the system (i.e. approx. <0.05 second!). Hence, in order to allow H2 to be

collected from the process, it is an important yet challenging task to avoid H2 consumption by

the H2-consuming methanogen in the system (by inhibiting the activity of methanogen) while at

the same time maintaining a low H2 partial pressure in order to remove the thermodynamic

bottleneck of the H2 producing reaction.

5. Conclusion

Undoubtedly, anaerobic fermentation of organic substrates would allow energy recovery

in the form of H2 gas. However, in terms of conversion efficiency the process releases only

about 17% of H2 in the substrate (i.e. glucose). Most of the H2 still remain in the end products

which resist for further H2 conversion due to microbiological and thermodynamical limitations.

Appendix 2

- 193 -

Appendix 2 (Supplementary information to Chapter 2, Section 2.3.1)

Estimation of Internal Resistance of the MFC

In Chapter 2 Table 1, the internal resistance (Rint) values are obtained by using the

polarization slope method (Logan 2008).

In the literature, there are several well accepted methods for the determination of Rint in a

MFC. These methods include electrochemical impedance spectroscopy (EIS) based on Nyquist

plots; current interrupt methods; power density peak and polarization slope method. The first

two methods require the use of a potentiostat that is able to varying the frequency of the

sinusoidal signal over a wide frequency range (typically from 10-4

to 106 Hz). The circuit

interruption method also requires the use of a potentiostat. This method also requires a very

effective and fast voltage measurement after interrupting the current.

Figure A2.1. Characteristics of a MFC polarization curve showing regions different types of

losses (overpotentials). In Region 2, the constant linear voltage drop allows an estimation

of internal resistance (Rint) of the MFC. Eo‘

cell = maximal electromotive force of the

redox reaction couples.

Due to the unavailability of suitable instruments for EIS and current interruption

methods, a relatively simple method, the polarization slope method was used to estimate the Rint

Current

Cell

Volta

ge

Eo’cell

Cell Voltage = OCV* - IRint

Re

gio

n 1

: A

cti

va

tio

n lo

ss

es

Re

gio

n 2

: O

hm

iclo

ss

es

Re

gio

n 3

: C

on

ce

ntr

ati

on

los

ses

Open Circuit Voltage (OCV)

Appendix 2

- 194 -

of the described MFC in Chapter 2. The Rint of the MFC can be directly obtained from the slope

of the linear portion of the polarization curve. As depicted Figure A2.1, region 2. The Rint values

of the MFC recorded at different times during the acclimation period are shown in Figure A2.2.

Figure A2.2. Determination of MFC internal resistances at different times during the

acclimation period using the polarization slope method. The slopes of the linear

regression curves are the internal resistance value (Ω) of the respective system. Data

obtained from Figure 2.3.

y = -2.26x + 761.21

R2 = 1.00

0

200

400

600

800

0 50 100

y = -1.20x + 733.12

R2 = 1.00

0

200

400

600

800

0 50 100 150 200

y = -0.97x + 711.45

R2 = 0.99

0

200

400

600

800

0 50 100 150 200

y = -0.54x + 696.85

R2 = 0.99

0

200

400

600

800

0 50 100 150 200 250 300 350 400 450 500

Current (mA)

Cell

Voltage (

mV

)

Day 5 Day 14

Day 22 Day 30

Appendix 3

- 195 -

Appendix 3 (Supplementary information to Chapter 4, Section 4.3.2)

In Chapter 4, application of computer feedback in BES processes is introduced with an

example of running a cyclic voltammetry analysis by feedback controlling the external

resistance of a MFC. Yet, this is only an example of using the proposed method for evaluating

electrochemical behavior of an anodophilic biofilm in a MFC. Other practical applications could

be possible based on a similar concept. In this appendix, another practical example is given:

MFC are designed to generate electrical power. To operate a MFC at its maximal power

generating capacity requires the knowledge of the optimum (or suitable) external resistance (of

the power consumer)(Aelterman et al. 2008). This knowledge can be obtained by using

computer feedback control of the external resistance as well. However, instead of a cyclic

potentiostatic control of the electrode potential as described in Chapter 4, a steady state control

of different anodic potential set points would allow the MFC to exhibit steady state maximal

power output (here ranged from 225 to 250 W m-3

) and the corresponding external resistance

setting (here ranged from 7 to 8 Ω) (Figure A3.1 A, C).

With this approach, one could operate the MFC at this selected external resistance to

sustain a maximal power output of the MFC by using computer feedback. In summary, using

computer feedback for BES process control should be given attention in future development of

the technology.

Appendix 3

- 196 -

Figure A3.1. Potentiostatic control of different anodic potential set points by feedback

controlling the external resistance of a MFC. (A) anodic potential and external resistance

vs. time; (B) current-time plot; (C) power density-time plot; (D) cathodic potential and

electrolyte Eh vs. time. Notes: Selected anodic set points: -400, -380, -350, -300, -275, -

250, -200, -150, -100, -50 mV vs. Ag/AgCl). Anolyte was saturated with the electron

donor substrate (here about 10 mM acetate) throughout the experiment. Red dotted

arrows indicate maximal power output and the corresponding anodic potential (here

between -350 and -375 mV vs. Ag/AgCl).

0

20

40

60

80

100

Curr

ent

(mA

)

-600

-450

-300

-150

0

150

0 100 200 300 400Time (min)

Po

ten

tia

l

(mV

vs.

Ag

/Ag

Cl)

.

CatholyteCathodeAnolyte

-450

-400

-350

-300

-250

-200

-150

-100

-50

An

od

ic P

ote

ntia

l

(mV

vs. A

g/A

gC

l).

1

10

100

1000

Exte

rna

l Re

sis

tan

ce

(Oh

m) .

Anodic Potential (Set point)

External resistance

0

50

100

150

200

250

Po

we

r D

en

sity (

W m

-3)

A.

B.

C.

D.

- 197 -

– END of the Thesis –

BES is a very exciting scientific research topic. However, towards a real-world

“Waste-to-Energy” application of BES there is still a long way to go!

We are just at the beginning of this long journey…

2010, Ka Yu

CV - K.Y.CHENG

- 198 -

Ka Yu CHENG

— Contributing to the sustainable development of humankind using innovative approaches —

Education

April 2006- 2009 Ph.D. (Candidate) in Environmental Engineering

Faculty of Sustainability, Environmental and Life Sciences

Murdoch University, Perth, Australia

Jul 2003 – Jul 2005 M.Phil. in Environmental Science

Research Topic: Phyto-remediation of PAH contaminated soil

Department of Biology, Faculty of Science

Hong Kong Baptist University (HKBU)

Sep 2000 – Sep 2003 B.Sc (Honor) in Applied Biology (Environmental Science)

Research Topic: Partitioning of PAHs in soil-water systems

Department of Biology, Faculty of Science

Hong Kong Baptist University (HKBU)

Awards/Achievements

2008 Huber Technology Prize (1st Prize)

The prize honors ideas, concepts and results of research that are an innovative contribution to

the reuse of energy and valuable materials from wastewater. (In IFAT 2008 - 15th International

Trade Fair for Water - Sewage - Refuse - Recycling, Munich, Germany)

2005 PhD Scholarship Awarded:

1. Murdoch International Fee Waiver Scholarship (Murdoch U) (Accepted)

2. Murdoch University Research Scholarship (Murdoch U) (Accepted)

3. Monash International Postgraduate Research Scholarship (Monash U)

4. Monash Graduate Scholarship (Monash U)

5. Griffith University Postgraduate Research Scholarship (Griffith U)

6. Australian Government International Postgraduate Research Scholarship (Griffith U)

2005 The Best Poster Award (1st Prize), International Symposium on Phytoremediation and

Ecosystem Health 2005 at Hangzhou, China

2003 President’s Honor Roll (HKBU)

2001 Ambassador, Beijing Environmental Protection Study Visit 2001, Environmental Awards

Scheme for University Students 2001

Peer-Review Activity

2010 – Present Reviewer for journals Electrochimica Acta, Water Research, Environmental Science

and Technology

2009 – Present Reviewer for journal Bioelectrochemistry

Patent

K. Y. Cheng, R. Cord-Ruwisch, Goen Ho (2009). A Wastewater Treatment Process. Australian

Provisional Patent.

CV - K.Y.CHENG

- 199 -

Publications

K. Y. Cheng, G. Ho and R. Cord-Ruwisch. (2010) An anodophilic biofilm catalyzes cathodic oxygen

reduction. Environmental Science and Technology. Vol. 44(1), 518-525.

K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2009) A new approach for in situ cyclic voltammetry of a

microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry. Vol. 74, 227-231

K. Y. Cheng, G. Ho and R. Cord-Ruwisch (2008) Affinity of microbial fuel cell biofilm for the anodic

potential. Environmental Science and Technology. Vol. 42(10), 3828-3834.

K. Y. Cheng, K. M. Lai, J. W. C. Wong. (2008) Effects of pig manure compost and nonionic-surfactant

Tween 80 on phenanthrene and pyrene removal in soil vegetated with Agropyron elongatum.

Chemosphere. Vol. 73, 791-797.

K. Y. Cheng, J. W. C. Wong. (2008). Fate of 14C-pyrene in soil-plant system amended with pig manure

compost and Tween 80: a growth chamber study. Bioresource Technology. Vol. 99, 8406-8412.

K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2007) Limitations of bio-hydrogen production by anaerobic

fermentation process: an overview. American Institute of Physics Conf. Proc. Vol. 941, 264-269.

K. Y. Cheng, J. W. C. Wong. (2006). Effect of synthetic surfactants on the solubilization and distribution

of PAHs in water/soil-water systems. Environmental Technology. Vol. 27, 835-844.

K. Y. Cheng, J. W. C. Wong. (2006). Combined effect of nonionic surfactant Tween 80 and DOM on the

behavior of PAHs in soil-water system. Chemosphere. Vol. 62, 1907-1916.

K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. (2004). Solubilization and desorption of PAHs in soil-aqueous

system by biosurfactants produced from Pseudomonas aeruginosa P-CG3 under thermophilic

condition. Environmental Technology. Vol. 25, 1159-1165.

K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. (2004). Effects of biosurfactants produced from Pseudomonas

aeruginosa P-CG3 on the solubilization and desorption of PAHs in soil-aqueous system under

thermophilic condition. Proceedings of the Fourth International Conference on Remediation of

Chlorinated and Recalcitrant Compounds, Monterey, California, USA; Publisher: Battelle Press,

Columbus, Ohio.

Conference Presentations

2008 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. A mixed anodophilic biofilm exhibits saturation

behavior with anodic potential in a microbial fuel cell. Microbial Fuel Cells First International

Symposium, Pennsylvania State University, 27-29 May 2008 at Pennsylvania State, USA (Poster)

2007 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. Computer-controlled microbial fuel cell enables

efficient electricity production from activated sludge. IWA Specialist Conference: 11th World

Congress on Anaerobic Digestion: Bioenergy for Our Future – Renewable Energy from Waste.

23-27 Sep 2007 at Brisbane, Queensland, Australia (Poster)

2007 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. Feasibility of bio-hydrogen production by anaerobic

fermentation process: A critical review. WREN International Conference Renewable Energy for

Sustainable Development in the Asia-Pacific Region. 4-8 Feb 2007 at Frementle, Perth, WA,

Australia (Oral)

2005 K. Y. Cheng and J. W. C. Wong. Phytoremediation of PAH contaminated soil using pig manure

compost and nonionic surfactant Tween 80. International Symposium on Phytoremediation and

Ecosystem Health. 10-13 Sep 2005 at Hangzhou, China (Best Poster Award)

2005 K. Y. Cheng and J. W. C. Wong*. Fate of 14C-pyrene in Soil-plant System Amended with Pig

Manure Compost and Tween 80: a Growth Chamber Study. International Symposium on

Phytoremediation and Ecosystem Health. 10-13 Sep 2005 at Hangzhou, China (*Oral)

2004 K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. Effects of biosurfactants produced from

Pseudomonas aeruginosa P-CG3 on the solubilization and desorption of PAHs in soil-aqueous

CV - K.Y.CHENG

- 200 -

system under thermophilic condition. Fourth International Conference on Remediation of

Chlorinated and Recalcitrant Compounds. 24-27 May 2004 at Monterey, California, US (Poster)

bioelectrochemical systems for energy ...―wastewater treatment process‖. australian provisional...

Documents